Information display method, information display device, and program

Abstract

An information display method comprising: acquiring a property prediction value of each of a plurality of compounds (step S112); acquiring first display method information indicating a method for displaying the property prediction value (step S113); generating a map indicating the property prediction value of each of the plurality of compounds in accordance with the first display method information (step S114); and generating and outputting an image including the map (step S115), the map having coordinate axes indicating at least two variables among a plurality of variables used for representing the structure of the compound.

Description

Information display method, information display device and program

[0001] The present disclosure relates to a technique for displaying information about compounds.

[0002] In the development of compound materials, in order to search for a composition formula or process conditions of a compound having desired properties, experiments have been conducted by changing the compounding ratio of raw materials, etc., to identify a composition formula or process conditions having good properties. For example, the following method has been disclosed to display the difference in properties when the composition formula or process conditions of a compound are changed.

[0003] Patent Document 1 discloses a method for displaying material properties on a ternary phase diagram when the composition of the composition formula is changed. Patent Document 2 discloses a method for displaying properties by arranging multiple data shown in pie charts in a matrix. Patent Document 3 discloses a method for inputting the blending ratios of a huge number of raw materials and outputting the blending ratios that have target properties. Patent Document 4 discloses a method for displaying multiple materials in two dimensions by changing markers based on prediction accuracy or material properties.

[0004] In recent years, technologies have been developed to predict material properties of arbitrary compositional formulas using predictors obtained by machine learning. For example, Non-Patent Document 1 discloses a method for predicting the thermodynamic stability of perovskite-based materials using a neural network. Non-Patent Document 2 discloses a method for predicting the probability that a compound will exhibit high ionic conductivity. Non-Patent Document 3 discloses a method for calculating the diffusion coefficient of ions. Non-Patent Document 4 discloses a method for optimizing material properties and creating phase diagrams using Bayesian optimization.

[0005] Japanese Patent No. 6632412 U.S. Patent No. 7199809 Japanese Patent No. 4009670 JP 2020-128962 A

[0006] Jonathan Schmidt et. al. “Predicting the Thermodynamic Stability of Solids Combining Density Functional Theory and Machine Learning” Chemistry of Materials, 29(12), 5090-5103 (2017) Austin D. Sendek et. al. “Holistic computational structure screening of more than 12000 candidates for solid lithium-ion conductor materials” Energy & Environmental Science, 10(1), 306-320 (2017) Xingfeng He et. al. “Statistical variables of diffusional properties from ab initio molecular dynamics simulations” npj Computational Materials, 4(1), 1-9 (2018) Kei Terayama, Koji Tsuda, Ryo Tamura. Efficient recommendation tool of materials by an executable file based on machine learning. Japanese Journal of Applied Physics, 58 (098001) 2019.

[0007] However, it is difficult for the techniques described in the above patent documents and non-patent documents to adequately support material development.

[0008] The present disclosure is intended to solve the above-mentioned problems, and provides an information display method and the like that can appropriately support material development.

[0009] In order to solve the above problem, an information display method according to one aspect of the present disclosure obtains predicted property values ​​for each of a plurality of compounds, obtains first display method information indicating a method for displaying the predicted property values, generates a map indicating the predicted property values ​​for each of the plurality of compounds according to the first display method information, and generates and outputs an image including the map, wherein the map has coordinate axes each indicating at least two variables out of a plurality of variables used to represent the structure of the compound.

[0010] These comprehensive or specific aspects may be realized as a system, an integrated circuit, or a computer-readable recording medium such as a CD-ROM, or may be realized as any combination of an apparatus, a system, a method, an integrated circuit, a computer program, and a recording medium. The recording medium may also be a non-transitory recording medium.

[0011] According to the present disclosure, material development can be appropriately supported.

[0012] Further advantages and benefits of certain aspects of the present disclosure will become apparent from the specification and drawings. Such advantages and / or benefits may be provided by some of the embodiments and features described in the specification and drawings, respectively, but not necessarily all of them may be provided to obtain one or more identical features.

[0013] FIG. 1 is a block diagram showing an example of the configuration of a display system according to Embodiment 1A. FIG. 2 is a diagram showing an example of a search range according to Embodiment 1A. FIG. 3 is a diagram showing an example of a composition formula represented by a combination of option data according to Embodiment 1A. FIG. 4 is a diagram showing an example of predicted property values ​​for each compound according to Embodiment 1A. FIG. 5A is a diagram showing an example of a map according to Embodiment 1A. FIG. 5B is a diagram showing an example of a map according to Embodiment 1A. FIG. 5C is a diagram showing an example of a map according to Embodiment 1A. FIG. 6 is a diagram showing another example of a map according to Embodiment 1A. FIG. 7A is a diagram showing another example of a map according to Embodiment 1A. FIG. 7B is a diagram showing another example of a map according to Embodiment 1A. FIG. 7C is a diagram showing another example of a map according to Embodiment 1A. FIG. 7D is a diagram showing another example of a map according to Embodiment 1A. FIG. 8 is a diagram showing another example of a map according to Embodiment 1A. FIG. 9A is a diagram showing another example of a map according to Embodiment 1A. FIG. 9B is a diagram showing another example of a map according to Embodiment 1A. FIG. 10 is a flowchart showing the processing operations of a display system according to Embodiment 1A. FIG. 11 is a diagram showing an example of a map according to Variation 1 of Embodiment 1A. FIG. 12 is a diagram showing an example of process variables in Modification 2 of Embodiment 1A. FIG. 13 is a block diagram showing an example of the configuration of a display system in Embodiment 1B. FIG. 14 is a diagram showing an example of experimental data stored in the experimental database in Embodiment 1B. FIG. 15 is a diagram showing an example of a map in Embodiment 1B. FIG. 16 is a diagram showing another example of a map in Embodiment 1B. FIG. 17 is a diagram showing another example of a map in Embodiment 1B. FIG. 18 is a diagram showing another example of a map in Embodiment 1B. FIG. 19 is a diagram showing another example of a map in Embodiment 1B. FIG. 20A is a diagram showing another example of a map in Embodiment 1B. FIG. 20B is a diagram showing another example of a map in Embodiment 1B. FIG. 21 is a diagram showing another example of a map in Embodiment 1B. FIG. 22A is a diagram showing another example of a map in Embodiment 1B.FIG. 22B is a diagram showing another example of a map in Embodiment 1B. FIG. 22C is a diagram showing another example of a map in Embodiment 1B. FIG. 23 is a flowchart showing processing operations of a display system in Embodiment 1B. FIG. 24A is a diagram showing an example of an image map in a modified example of Embodiment 1B. FIG. 24B is a diagram showing another example of an image map in a modified example of Embodiment 1B. FIG. 25 is a block diagram showing an example of the configuration of a display system in Embodiment 2A. FIG. 26 is a diagram showing an example of an image element map on which candidate points are superimposed in Embodiment 2A. FIG. 27 is a diagram showing another example of an image element map on which candidate points are superimposed in Embodiment 2A. FIG. 28 is a diagram showing another example of an image element map on which candidate points are superimposed in Embodiment 2A. FIG. 29 is a diagram showing another example of an image element map on which candidate points are superimposed in Embodiment 2A. FIG. 30 is a diagram showing another example of an image element map on which candidate points are superimposed in Embodiment 2A. FIG. 31 is a diagram showing another example of an image element map on which candidate points are superimposed in Embodiment 2A. FIG. 32 is a diagram showing another example of an image element map on which candidate points are superimposed in Embodiment 2A. FIG. 33 is a flowchart showing processing operations of a display system in Embodiment 2A. FIG. 34 is a diagram showing an example of an image map in a modified example of Embodiment 2A. FIG. 35 is a diagram showing another example of an image map in a modified example of Embodiment 2A. FIG. 36 is a diagram showing another example of an image map in a modified example of Embodiment 2A. FIG. 37 is a block diagram showing an example of the configuration of a display system in Embodiment 2B. FIG. 38 is a diagram showing an example of an image element map on which candidate points and experimental characteristic values ​​are superimposed in Embodiment 2B. FIG. 39 is a diagram showing another example of an image element map on which candidate points and experimental characteristic values ​​are superimposed in Embodiment 2B. FIG. 40 is a diagram showing an example of state transitions of an image element map in Embodiment 2B. FIG. 41 is a diagram showing another example of state transitions of an image element map in Embodiment 2B. FIG. 42 is a flowchart showing processing operations of a display system in Embodiment 2B. FIG. 43 is a block diagram showing an example of the configuration of a display system in a modified example of Embodiment 2B.FIG. 44 is a diagram showing an example of a history of characteristic display images in a modified example of Embodiment 2B. FIG. 45 is a block diagram showing an example of the configuration of a display system in Embodiment 2C. FIG. 46 is a diagram showing an example of candidate point data stored in a candidate point database in Embodiment 2C. FIG. 47 is a diagram showing an example of an image element map on which candidate points and characteristic experiment values ​​are superimposed in Embodiment 2C. FIG. 48 is a flowchart showing the processing operation of a display system in Embodiment 2C. FIG. 49 is a block diagram showing an example of the configuration of a display system in Embodiment 3A. FIG. 50 is a diagram showing an example of information stored in an evaluation display database in Embodiment 3A. FIG. 51 is a diagram showing an example of each data stored in an experiment database in Embodiment 3A. FIG. 52 is a diagram showing an example of compound basic data in Embodiment 3A. FIG. 53 is a diagram showing an example of compound detailed data in Embodiment 3A. FIG. 54 is a diagram showing an example of a map in Embodiment 3A. FIG. 55 is a diagram showing another example of a map in Embodiment 3A. FIG. 56 is a diagram showing a legend for the map in FIG. 55. FIG. 57A is a diagram for explaining an example of an image transition accompanying a change in calculation method information in Embodiment 3A. FIG. 57B is a diagram illustrating an example of an image transition accompanying a change in calculation method information in Embodiment 3A. FIG. 58A is a diagram illustrating an example of an image transition accompanying a change in search range information in Embodiment 3A. FIG. 58B is a diagram illustrating an example of an image transition accompanying a change in search range information in Embodiment 3A. FIG. 59A is a diagram illustrating an example of an image transition accompanying a change in first display target information in Embodiment 3A. FIG. 59B is a diagram illustrating an example of an image transition accompanying a change in first display target information in Embodiment 3A. FIG. 60A is a diagram illustrating an example of an image transition accompanying a change in display range information in Embodiment 3A. FIG. 60B is a diagram illustrating an example of an image transition accompanying a change in display range information in Embodiment 3A. FIG. 61A is a diagram illustrating an example of an image transition accompanying a change in search range information in Embodiment 3A.FIG. 61B is a diagram illustrating an example of an image transition accompanying a change in search range information in Embodiment 3A. FIG. 62 is a diagram illustrating an example of an image transition accompanying a change in map array information in Embodiment 3A. FIG. 63 is a flowchart illustrating the processing operation of a display system in Embodiment 3A. FIG. 64 is a diagram illustrating an example of a first image in a modified example of Embodiment 3A. FIG. 65 is a diagram illustrating another example of a first image in a modified example of Embodiment 3A. FIG. 66 is a flowchart illustrating a detailed example of generation of a first image by an image processing unit in a modified example of Embodiment 3A. FIG. 67 is a block diagram illustrating an example of the configuration of a display system in Embodiment 3B. FIG. 68 is a diagram illustrating an example of information stored in the experiment display database in Embodiment 3B. FIG. 69A is a diagram illustrating an example of an image transition accompanying a change in second display target information in Embodiment 3B. FIG. 69B is a diagram illustrating an example of an image transition accompanying a change in second display target information in Embodiment 3B. FIG. 70A is a diagram illustrating an example of an image transition accompanying a change in display condition information in Embodiment 3B. FIG. 70B is a diagram illustrating an example of an image transition accompanying a change in display condition information in Embodiment 3B. FIG. 71 is a flowchart showing the processing operation of the display system in embodiment 3B. FIG. 72 is a block diagram showing an example of the configuration of the display system in embodiment 3C. FIG. 73A is a diagram for explaining an example of an image transition accompanying a change in date setting information in embodiment 3C. FIG. 73B is a diagram for explaining an example of an image transition accompanying a change in date setting information in embodiment 3C. FIG. 74 is a flowchart showing the processing operation of the display system in embodiment 3C. FIG. 75 is a block diagram showing an example of the configuration of a display system in embodiment 3D. FIG. 76 is a diagram showing an example of information stored in the evaluation display database in embodiment 3D. FIG. 77 is a diagram showing an example of information stored in the experiment display database in embodiment 3D. FIG. 78 is a diagram showing an example of an image transition accompanying the acquisition of position information in embodiment 3D.Fig. 79 is a diagram showing an example of a composition image showing a composition formula including element coefficients as non-visualized variables in embodiment 3D. Fig. 80 is a diagram showing an example of a composition image in the case where a process variable is used as a non-visualized variable in embodiment 3D. Fig. 81 is a flowchart showing the processing operation of the display system in embodiment 3D.

[0014] (Knowledge that led to the present disclosure) In recent years, in fields such as image recognition and natural language processing, recognition and identification methods using machine learning and the like have made significant progress, and are beginning to be applied to predicting material properties.

[0015] For example, Non-Patent Document 1 discloses a method for predicting the thermodynamic stability of perovskite-based materials using a neural network. Perovskite-based materials are materials expressed by the composition formula ABX_3, where A, B, and X are designated as different elements. According to the method of Non-Patent Document 1, it is possible to predict the thermodynamic stability of a composition formula in which A, B, and X are each assigned an unexplored combination of elements. This method can reduce the number of labor-intensive experiments and promote material development.

[0016] However, although the technology disclosed in Non-Patent Document 1 is capable of outputting a huge number of prediction results, each prediction result is presented individually, making it difficult to recognize the relationship between each prediction result.

[0017] Furthermore, Non-Patent Document 4 discloses a method for optimizing material properties and creating a phase diagram using Bayesian optimization, as described above. As a demonstration example, a low-melting-point composition search in a NaF-KF-LiF system or experimental values ​​are displayed superimposed on a triangular phase diagram as a phase diagram search result, along with predicted values. However, the method does not have a function for changing the display method.

[0018] Patent Document 1 discloses a method for displaying the properties of a compound when the composition of a composition formula is changed on a triangular phase diagram. However, simply displaying a triangular phase diagram as in Patent Document 1 only presents the compounding ratio of three elements, i.e., the properties of a material within a range represented by two variables. Therefore, when there are many types of search variables for a compound, it is not possible to display the entire composition range of the compound to be searched, making it difficult to recognize the relationship between the properties of each compound.

[0019] Patent Document 2 discloses a method for visualizing characteristics by arranging multiple pieces of data shown in pie charts in a matrix. However, with the technology disclosed in Patent Document 2, in the case of compounds with a large number of search variables, such as perovskite-based materials, it is not possible to display the entire range of the compounds to be searched for, making it difficult to recognize the relationships between the properties of the compounds.

[0020] Patent Document 3 discloses a method for inputting the compounding ratios of a huge number of raw materials and outputting the compounding ratios having the target properties. However, with the technology disclosed in Patent Document 3, the output results are presented locally, making it impossible to display the entire composition range of the compound being searched for, making it difficult to recognize the relationship between the properties of each compound. Furthermore, while the markers are changed and displayed depending on the compounding ratios, the technology does not disclose a method for changing the markers based on variables not used in the display when the composition formula is expressed by multiple variables, making it difficult to grasp the information of the variables not used in the display.

[0021] Patent Document 4 discloses a method for two-dimensionally displaying multiple materials by changing markers based on prediction accuracy and material properties. However, this method only presents the properties of the materials within the search variables compressed into two dimensions. Therefore, when a compound has many search variables, it is difficult to recognize the overall properties of the compound. Furthermore, it does not disclose a method for changing the display method depending on search variables exceeding two variables. In other words, when there are three or more search variables, it is not possible to display predicted and experimental values ​​across the entire range of the compound being searched, making it difficult to recognize the relationship between each of the compound's properties.

[0022] As described above, the methods used in the above patent and non-patent documents lack or do not adequately provide information useful for material development, resulting in insufficient support for material development.

[0023] Therefore, in an information display method according to one aspect of the present disclosure, as shown in, for example, Embodiments 1A and 1B, predicted property values ​​for each of a plurality of compounds are acquired, first display method information indicating a display method for the predicted property values ​​is acquired, a map indicating the predicted property values ​​for each of the plurality of compounds is generated according to the first display method information, and an image including the map is generated and output, the map having coordinate axes indicating at least two variables among a plurality of variables used to represent the compound structure. For example, the image is output and displayed on a display unit. It can also be said that the plurality of variables determine, for example, the compound structure. The information display method according to the present disclosure can also be said to be a characteristic display method.

[0024] As a result, first display method information is acquired, and a map showing predicted property values ​​for each of the multiple compounds is generated according to the display method indicated by the first display method information. As a result, if the display method indicated by the first display method information is set in accordance with the purpose of searching for new materials, displaying an image including the generated map can appropriately support the search for new materials, i.e., material development. Furthermore, because the map has the above-mentioned coordinate axes, the predicted property values ​​of each of the multiple compounds can be displayed at positions corresponding to the respective structures of the compounds. Furthermore, because the predicted property values ​​of each of the many compounds to be searched for are displayed as a map, the predicted property values ​​can be appropriately visualized, allowing the user to easily recognize the relationship between the compound structures and the predicted property values. In other words, the user can easily grasp the overall picture of the predicted property values ​​in the map. The search for new materials is also referred to as material search.

[0025] Alternatively, an information display method according to one aspect of the present disclosure includes acquiring predicted property values ​​for each of a plurality of compounds, and outputting an image including a map generated using the acquired predicted property values, the map having coordinate axes each representing at least two variables used to express the structure of the compounds, and showing the predicted property values ​​for each of the plurality of compounds. Even with such an information display method, the same effects as those described above can be obtained.

[0026] Furthermore, the first display method information may indicate at least one of a method for determining a color for indicating the predicted property value and a method for determining a display format for indicating the predicted property value as a display method for the predicted property value. For example, a color corresponding to the predicted property value is determined by the color determination method, and a color shading corresponding to the predicted property value is determined by the display format determination method. As a result, in generating the map, the map may be generated that indicates, for each of the plurality of compounds, a color or a color shading corresponding to the predicted property value of the compound.

[0027] This allows the user to easily grasp the magnitude of the predicted property value, for example, the difference in the predicted property values ​​between compounds with similar structures, because the predicted property value of each compound is displayed in a color or shade of color corresponding to the predicted property value.

[0028] Furthermore, in generating the map, the predicted characteristic values ​​may be discretized, and the map indicating the discretized predicted characteristic values ​​may be generated.

[0029] This improves the visibility of areas on the map that have similar predicted characteristic values, making it easier to understand the change trend or distribution of predicted characteristic values, and reducing the amount of work required to understand predicted characteristic values.

[0030] The first display method information may indicate, as a display method of the predicted characteristic values, at least one of: (a) displaying on the map predicted characteristic values ​​that are equal to the reference value in a predetermined color or shade of a color; (b) displaying predicted characteristic values ​​that are greater than the reference value in a first series of colors and displayed predicted characteristic values ​​that are smaller than the reference value in a second series of colors; (c) superimposing a boundary line on the map between a first region in which a predicted characteristic value that is equal to the reference value is displayed and a second region in which a predicted characteristic value different from the reference value is displayed; (d) displaying, in a predetermined color or shade of a color, one of a third region in which a predicted characteristic value that satisfies a predetermined condition is displayed and a fourth region in which a predicted characteristic value that does not satisfy the predetermined condition is displayed; and (e) superimposing a pattern of a plurality of dots or stripes on one of the third region and the fourth region.

[0031] This allows the user to easily grasp the areas where the characteristic prediction value is equal to the reference value, the areas where the characteristic prediction value is greater than or less than the reference value, the boundaries between these areas, or the areas where the specified conditions are met or not met.

[0032] The reference value may be the average or median of the predicted property values ​​of each of the plurality of compounds, or a value designated by the user.

[0033] This makes it easy to set appropriate reference values ​​for dividing the entire map into multiple regions or displaying the boundaries between those regions.

[0034] Furthermore, in generating the map, a gradient of the predicted characteristic value shown on the map may be specified, and an arrow indicating the direction and magnitude of the gradient may be superimposed on the map.

[0035] This allows the user to easily grasp the changes in the predicted characteristic value for at least two variables by looking at the arrows on the map.

[0036] The information display method may further include acquiring a plurality of variables used to express the structure of the compound and a plurality of option data items indicating, for each of the plurality of variables, values ​​or elements that the variable can take, and acquiring the predicted property value for each combination of option data items obtained by selecting one option data item from the plurality of option data items for each of the plurality of variables. For example, acquiring the predicted property value for a compound having a structure corresponding to each combination may include acquiring, for each combination, the predicted property value for a compound having a structure corresponding to the combination using a predetermined algorithm.

[0037] This allows for appropriate predicted property values ​​to be obtained. That is, the property values ​​of multiple compounds with different compositions can be appropriately predicted. Furthermore, when the predetermined algorithm is updated based on experiments on compounds that are conducted from time to time, the prediction accuracy of the property values ​​can be improved.

[0038] Furthermore, if there is an unused variable among the plurality of variables that is a variable other than the at least two variables used on the coordinate axes of the map, the first display method information may indicate that the characteristic prediction value is displayed using the unused variable as a display method for the characteristic prediction value.

[0039] This allows predicted property values ​​to be displayed using not only the variables used on the coordinate axes of the map but also non-utilized variables, thereby providing appropriate support for material development.

[0040] Furthermore, the first display method information indicates that, when there is an unutilized variable that is a variable other than the at least two variables used for the coordinate axes of the map among the plurality of variables, a first value, a second value, or a numerical value within a predetermined numerical range is substituted for the unutilized variable as a display method of the predicted property values; (a) when a first value is substituted for the unutilized variable, the map generation generates the map indicating the predicted property values ​​of each of the plurality of compounds having a configuration expressed using the unutilized variable indicating the first value designated by a user; and (b) when a second value is substituted for the unutilized variable, the map generation determines the second value so that the predicted property values ​​shown in the map satisfy a predetermined condition. and generating the map showing predicted property values ​​of each of the plurality of compounds having a configuration expressed using the non-utilized variables indicating the determined second value; (c) when each numerical value within the predetermined numerical range is substituted for the non-utilized variables, the map is generated by calculating, for each position on the map, an average value of the predicted property values ​​of the plurality of compounds having a configuration expressed using the at least two variables each indicating a numerical value corresponding to the position; and generating the map showing, for each position on the map, the average value of the predicted property values ​​calculated for that position, wherein the non-utilized variables of the plurality of compounds for which the average predicted property values ​​are calculated may indicate mutually different numerical values ​​within the predetermined numerical range.

[0041] As a result, in the case of (a), the user specifies a first value for the non-utilized variable, allowing the user to arbitrarily select the configuration of the compound having the predicted property value to be displayed on the map. Furthermore, in the case of (b), by setting predetermined conditions as the conditions required for material search, a map that satisfies the conditions required for material search can be easily displayed without the user having to specify values ​​for the non-utilized variable. As a result, the efficiency of material search can be improved. In other words, material development can be appropriately supported. Furthermore, in the case of (c), changes in the values ​​that the non-utilized variable can take can be reflected in the predicted property value on the map, thereby improving the robustness of the predicted property value with respect to the non-utilized variable.

[0042] In addition, in obtaining the predicted property values, the predicted property values ​​of each of the plurality of compounds may be obtained from at least one predictor for predicting property values ​​of compounds, which is stored in a predictor database.

[0043] This makes it possible to obtain appropriate predicted property values ​​for each of a plurality of compounds.

[0044] Furthermore, the information display method may further include acquiring experimental characteristic values ​​of each of one or more compounds that have been tested, and in generating the image, superimposing the experimental characteristic values ​​of each of the one or more compounds that have been tested at positions on a map that correspond to the structure of the compound, and generating the image including the map on which the experimental characteristic values ​​are superimposed.

[0045] This allows easy comparison of predicted property values ​​with experimental property values ​​within the range shown on the map, thereby improving the efficiency of material search.

[0046] In addition, the information display method may further acquire second display method information indicating a display method for the characteristic experimental value, and in generating the image, the characteristic experimental value may be superimposed on the map according to the second display method information.

[0047] As a result, the experimental characteristic values ​​are superimposed on the map in accordance with the display method indicated by the second display method information, and therefore the experimental characteristic values ​​can be displayed in a manner that suits the user's purpose in material exploration, depending on the display method setting, thereby improving the efficiency of material exploration and appropriately supporting material development.

[0048] Furthermore, when the predicted characteristic value is shown on the map in a first display mode that is a color or a shade of color corresponding to the predicted characteristic value, and the experimental characteristic value is superimposed on the map as a mark having a second display mode that is a color or a shade of color corresponding to the experimental characteristic value, the second display method information may indicate, as a display method for the experimental characteristic value, that the scale of the second display mode for the experimental characteristic value is made to match the scale of the first display mode for the experimental characteristic value.

[0049] This allows the predicted characteristic values ​​and the experimental characteristic values ​​to be on the same scale, making it easier to compare the predicted characteristic values ​​with the experimental characteristic values.

[0050] The second display method information may also indicate, as a display method for the characteristic experimental values, a rule that specifies the order of overlapping of the marks when multiple characteristic experimental values ​​are superimposed on the map in the form of overlapping marks, and in generating the image, the marks for the multiple characteristic experimental values ​​are superimposed on the map in the form of overlapping marks according to the rule. The rule may also specify that (a) the larger the characteristic experimental value, the closer the mark for the characteristic experimental value is to be placed to the front side, (b) the closer the characteristic experimental value is to a predetermined value, the closer the mark for the characteristic experimental value is to be placed to the front side, or (c) the closer the characteristic experimental value is to the characteristic predicted value indicated at the position on the map where the characteristic experimental value is superimposed, the closer the mark for the characteristic experimental value is to be placed to the front side.

[0051] This prevents marks indicating good experimental characteristic values, such as large experimental characteristic values, experimental characteristic values ​​close to a predetermined value, or experimental characteristic values ​​close to a predicted value, from being obscured by marks indicating other experimental characteristic values, thereby improving the efficiency of material search.

[0052] Furthermore, the second display method information may indicate, as a display method for the characteristic experimental value, when there is no position on the map corresponding to the configuration of the compound having the characteristic experimental value, (a) superimposing the characteristic experimental value at a position on the map corresponding to a configuration that is closest to the configuration of the compound having the characteristic experimental value, or (b) not superimposing the characteristic experimental value on the map, and in generating the image, processing for superimposing the characteristic experimental value on the map may be performed in accordance with the second display method information.

[0053] As a result, in the case of (a), even if the experimental characteristic value of a compound having a structure that does not correspond to any position on the map, the experimental characteristic value is superimposed on the position corresponding to the structure closest to the structure. Therefore, the experimental characteristic value of a compound having a structure that does not correspond to the map is also displayed in appropriate association with the map, thereby improving the efficiency of material search by the user. Furthermore, in the case of (b), the experimental characteristic value that does not correspond to the map is not superimposed, thereby reducing misunderstandings that may arise from the superimposition of the experimental characteristic value.

[0054] Furthermore, the second display method information may indicate, as a display method for the characteristic experimental values, that, among the characteristic experimental values ​​of the one or more compounds that have been tested, characteristic experimental values ​​that are equal to or greater than a predetermined first threshold value are superimposed on the map, and characteristic experimental values ​​that are less than the first threshold value are not superimposed on the map.

[0055] This allows, for example, when many experimental property values ​​are obtained, to narrow down the obtained experimental property values ​​to only important ones and superimpose them on the map, without superimposing unimportant experimental property values ​​that are treated as noise on the map, making it possible to make the important experimental property values ​​superimposed on the map easier to see, thereby improving the efficiency of material search.

[0056] Furthermore, the second display method information indicates, as a method of displaying the characteristic experimental values, that characteristic experimental values ​​that satisfy predetermined conditions among the characteristic experimental values ​​of the one or more compounds that have been tested are superimposed on the map in a manner that emphasizes them more than characteristic experimental values ​​that do not satisfy the predetermined conditions, and in generating the image, the characteristic experimental values ​​that satisfy the predetermined conditions may be superimposed on the map in a manner that emphasizes them in accordance with the second display method information.

[0057] This allows the user to visually and easily determine whether each property experiment value displayed on the map satisfies a predetermined condition, thereby improving the efficiency of material search.

[0058] The predetermined condition may be (a) that the experimental characteristic value is a characteristic experimental value obtained within a predetermined period of time from the present, (b) that the experimental characteristic value is one of a predetermined number of experimental characteristic values ​​obtained most recently, (c) that the experimental characteristic value is equal to or greater than a predetermined second threshold, or (d) that the difference between the experimental characteristic value and a predicted characteristic value obtained for a compound having the same structure as the compound having the experimental characteristic value is equal to or greater than a predetermined third threshold or less than the third threshold.

[0059] As a result, in (a) and (b), the user can visually and easily grasp new experimental property values. In addition, in (c), the user can visually and easily grasp, for example, important experimental property values. In (d), the user can visually and easily grasp experimental property values ​​that are close to predicted property values. As a result, the efficiency of material search can be improved, and material development can be appropriately supported.

[0060] Furthermore, in the information display method, when at least one non-utilized variable is associated with each of the characteristic experimental values ​​of the one or more tested compounds, the product of the number of cases of the at least one non-utilized variable may be calculated as the number of types of mark shapes, and in generating the image, for each of the characteristic experimental values ​​of the one or more tested compounds, a shape from the number of types corresponding to a combination of data indicated by each of the at least one non-utilized variable associated with the characteristic experimental value may be selected, and the mark of the selected shape may be superimposed on the map.

[0061] This allows the user to visually and easily grasp the data of at least one non-utilized variable associated with each experimental characteristic value superimposed on the map from the shape of the mark for that experimental characteristic value, thereby improving the efficiency of materials search and appropriately supporting materials development.

[0062] Furthermore, each of the at least one unutilized variable may represent, as the data, process conditions used to produce a compound having the experimental property value, or an attribute of the compound having the experimental property value.

[0063] This allows the user to visually and easily grasp the process conditions or attributes of a compound having an experimental characteristic value displayed superimposed on the map from the shape of the mark for that experimental characteristic value.

[0064] In addition, in calculating the number of types, the number of types of the mark form may be calculated as a number between 2 and 15 inclusive.

[0065] For example, if the number of types of mark shapes were 16 or more, it would be difficult to distinguish between the shapes. Therefore, by limiting the number of types to 2 or more and 15 or less, the distinguishability between the shapes can be improved. As a result, the user can more easily grasp the data for at least one non-utilized variable associated with each characteristic experimental value superimposed on the map. This can further improve the efficiency of material search.

[0066] Furthermore, the attribute of the compound may be (a) the type of crystalline phase of the compound, (b) the type of raw material of the compound, or (c) whether or not the raw material of the compound remained when the compound was produced.

[0067] This allows the user to visually and easily grasp the type of crystalline phase, the type of raw material, or whether or not the raw material remains of a compound having a characteristic experimental value superimposed on the map from the shape of the mark for that characteristic experimental value.

[0068] Furthermore, the map may include a plurality of image element maps arranged in a matrix along a first coordinate axis and a second coordinate axis, each of the plurality of image element maps having a third coordinate axis and a fourth coordinate axis, and in generating the map, the first coordinate axis, the second coordinate axis, the third coordinate axis, and the fourth coordinate axis may be associated with a first variable, a second variable, a third variable, and a fourth variable of the plurality of variables, respectively, and for each of the plurality of compounds, an image element map associated with a value of the first variable and a value of the second variable used to represent a configuration of the compound may be identified from the plurality of image element maps, and a predicted property value of the compound may be mapped to a position on the identified image element map corresponding to the value of the third variable and the value of the fourth variable used to represent the configuration of the compound.

[0069] This allows the map to show predicted property values ​​for multiple compound structures, each expressed by four variables, and to clearly display predicted property values ​​for compounds over a wide range, thereby improving the efficiency of materials search and appropriately supporting materials development.

[0070] Furthermore, in generating the image, if there is no image element map among the plurality of image element maps for each of the one or more experimental compounds that is associated with the value of the first variable and the value of the second variable used to represent the configuration of the compound, image element maps associated with values ​​closest to the values ​​of the first variable and the second variable may be identified instead of the image element map, and the characteristic experimental value of the compound may be superimposed on the identified image element map at a position corresponding to the value of the third variable and the value of the fourth variable used to represent the configuration of the compound.

[0071] For example, since the first variable and the second variable are discrete variables, there may be cases where there are no image element maps associated with the values ​​of the first variable and the second variable. However, in this embodiment, the image element maps associated with the values ​​closest to those variables are identified, and the characteristic experimental value is superimposed on that image element map. Therefore, it is possible to prevent the characteristic experimental value from being hidden from the image element map due to the discrete variables.

[0072] In addition, in acquiring the first display method information, the first display method information generated in response to an input operation by a user may be acquired.

[0073] This allows the user to arbitrarily set the display method for the predicted characteristic values, thereby improving convenience.

[0074] Furthermore, the acquisition of the first display method information may involve determining a display method for the predicted property values ​​based on the acquired predicted property values ​​for each of the plurality of compounds. For example, the acquisition of the first display method information may involve determining, as a display method for the predicted property values, that the map be generated by associating predetermined colors or shades of colors for indicating maximum and minimum values ​​of the predicted property values ​​on the map with the maximum and minimum values ​​of the acquired predicted property values ​​for the plurality of compounds, respectively.

[0075] As a result, the display method for the predicted property value is determined based on the predicted property value, so that a display suited to the predicted property value can be performed, thereby improving the efficiency of material search.

[0076] Furthermore, the information display method further includes acquiring position information indicating a position on the map of the predicted property value or the experimental property value, and generating the image by acquiring compositional formula data related to a compositional formula of a compound corresponding to the position indicated by the position information, and superimposing a composition image indicating the compositional formula data on the map, the compositional formula data including a non-utilized variable associated with the compound having the predicted property value or the experimental property value, and the non-utilized variable may be a variable of the plurality of variables other than the at least two variables used on the coordinate axes of the map.

[0077] As a result, even if it is not possible to determine the non-utilized variables of the compound corresponding to the position from the position of the predicted property value or the experimental property value, the composition image including the non-utilized variables is superimposed on the map, and therefore the non-utilized variables can be easily determined by looking at the displayed composition image.

[0078] Each process or operation included in the information display method is executed by an information display device or a computer. For example, the information display device or the computer includes a processor and a computer-readable non-transitory recording medium. The recording medium is, for example, a memory, and stores a program for causing the processor to execute each process or operation included in the information display method. The processor reads and executes the program from the recording medium. This executes each process or operation included in the information display method.

[0079] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. Each of the embodiments described below represents a preferred specific example of the present disclosure. Therefore, the numerical values, shapes, materials, components, arrangement positions and connection forms of the components shown in the following embodiments are merely examples and are not intended to limit the present disclosure. Therefore, among the components in the following embodiments, components that are not described in the independent claims that represent the highest concept will be described as optional components.

[0080] It should be noted that the figures are schematic diagrams and are not necessarily precise illustrations. In addition, in the figures, substantially identical components are denoted by the same reference numerals, and redundant explanations are omitted or simplified. The effects of the above-described information display method are also realized in an information display device and a program.

[0081] The present disclosure also includes embodiments 1, 2, and 3, with embodiment 1 including embodiments 1A and 1B. Embodiment 2 includes embodiments 2A, 2B, and 2C, and embodiment 3 includes embodiments 3A, 3B, 3C, and 3D. Each embodiment will be described in turn below.

[0082] (Embodiment 1A) A display system according to this embodiment displays predicted property values ​​of each of a plurality of compounds in the form of a map.

[0083] [Configuration of Display System 100] Fig. 1 is a block diagram showing an example of the configuration of a display system 100 according to the present embodiment. The display system 100 shown in Fig. 1 includes an input unit 110, a predictor database (DB) 120, a characteristic display device 130, and a display unit 140. The characteristic display device 130 is an example of an information display device.

[0084] [Input Unit 110] The input unit 110 accepts an input operation by a user and outputs an input signal corresponding to the input operation to the characteristic display device 130. For example, the input unit 110 may be configured as a keyboard, a touch sensor, a touch pad, a mouse, or the like.

[0085] [Predictor Database 120] The predictor database 120 is a recording medium that stores a predictor, which is, for example, a computer program. Specifically, the predictor database 120 stores a predictor for predicting the respective property values ​​of multiple compounds. The predicted property values ​​are, for example, electrical conductivity, thermal conductivity, band gap, etc., and are hereinafter also referred to as predicted property values. The predictor is also a computer program based on a predetermined calculation algorithm. The predetermined calculation algorithm may be an algorithm obtained by machine learning. The predictor database 120 is composed of, for example, a non-volatile memory. The predictor outputs the property values ​​of a compound in response to input data related to the compound. The data is, for example, the compositional formula of the compound.

[0086] The predictor may be of any type as long as it can output a predicted property value of a compound. The predicted property value output from the predictor may be a value calculated using the predictors described in Non-Patent Document 1 and Non-Patent Document 2, for example. Non-Patent Document 1 discloses a method for predicting the thermodynamic stability of perovskite-based compounds. Non-Patent Document 2 discloses a method for predicting the probability that a compound will exhibit high ionic conductivity. Note that the term "compound" is synonymous with "material."

[0087] Furthermore, the predicted property values ​​output from the predictor may be, for example, property values ​​calculated by density functional theory. A crystal structure must be specified to perform density functional theory processing. The crystal structure may be specified by an input signal from the input unit 110. Even if the crystal structure of the compound to be predicted is unknown, the predictor may output the property values ​​of a compound that has the same compositional formula as the compound and a known crystal structure as the predicted property value of the compound to be predicted. Furthermore, if the compositional formula of the compound to be predicted differs from the compositional formula of a predictable compound, the predictor may create a provisional crystal structure by replacing some elements in the compositional formula of the compound to be predicted with other elements, and then perform density functional theory calculations on the compound having the created crystal structure. Alternatively, if the crystal structure of one of the compounds is experimentally verified, the predictor may adopt that crystal structure and perform density functional theory calculations in the same manner as described above.

[0088] Furthermore, the predicted property value output from the predictor may be a property value calculated using a technique called Ab initio molecular dynamics (AIMD) described in Non-Patent Document 3. Non-Patent Document 3 discloses a method for calculating the diffusion coefficient of ions. Note that the property value is not limited to a predicted value, and may be an experimentally obtained property value such as electrical conductivity or thermal conductivity.

[0089] The predictor database 120 outputs to the predicted value acquisition unit 132 a predictor that meets the request of the predicted value acquisition unit 132, that is, a predictor that is suitable for predicting the properties of a compound.

[0090] [Display Unit 140] The display unit 140 acquires an image from the characteristic display device 130 and displays the image. Such a display unit 140 is, for example, a liquid crystal display, a plasma display, an organic EL (Electro-Luminescence) display, or the like, but is not limited to these.

[0091] The display unit 140 is, for example, a display unit provided in an information terminal, an information terminal equipped with an electronic experiment notebook, etc. The display unit 140 may also be a display unit provided in a synthesis apparatus used to synthesize raw materials or compounds, a reaction apparatus used to react raw materials or compounds, an analysis apparatus used to analyze compounds, an evaluation apparatus used to evaluate compounds, etc.

[0092] [Property display device 130] The property display device 130 receives an input signal from the input unit 110 and predicts property values ​​of multiple compounds corresponding to the input signal using a predictor in the predictor database 120. Furthermore, the property display device 130 receives an input signal indicating a display method from the input unit 110 as first display method information, and generates an image indicating the predicted property values ​​of each compound as predicted property values ​​according to the display method. The property display device 130 then outputs the generated image to the display unit 140.

[0093] [Configuration of characteristic display device 130] The characteristic display device 130 includes a search range acquisition unit 131, a predicted value acquisition unit 132, a display method acquisition unit 133, and an image processing unit 134. The characteristic display device 130 may be configured with a processor such as a central processing unit (CPU) and a memory. In this case, the processor functions as the characteristic display device 130 by executing a computer program stored in the memory, for example. The memory may be volatile or non-volatile, or may be configured with a volatile memory and a non-volatile memory.

[0094] [Search Range Acquisition Unit 131] The search range acquisition unit 131 acquires an input signal indicating a search range for compounds from the input unit 110. Then, the search range acquisition unit 131 outputs the input signal as a search range signal to the predicted value acquisition unit 132. The search range for compounds is the range of the composition formula, structure, etc. of each compound whose characteristic value is to be predicted. In this embodiment, the compound to be predicted is a solid electrolyte, and the composition formula of the compound is expressed, for example, by the following (Formula 1).

[0095] Li ２－３ａ－４ｂ (M3１－ｘ M3' ｘ ) a(M4 １－ｙ M4' ｙ ) １＋ｂ O ３ ...(Formula 1)

[0096] The search range acquisition unit 131 acquires each piece of data that can be taken by the variables a, b, x, y, M3, M3', M4, and M4' in (Equation 1) (i.e., option data, described later) as the search range.

[0097] 2 is a diagram showing an example of a search range. Search points included in the search range are expressed by one or more pre-specified data (for example, elements Li and O and their coefficients) and one or more range variables, and when values ​​or elements are specified for each of all range variables, one composition formula is shown. In other words, when values ​​or elements are specified for each of all range variables, the data expression shown using those range variables shows one structure of a compound, i.e., one composition formula. The data expression can be, for example, the Li ２－３ａ－４ｂ (M3 １－ｘ M3' ｘ ) ａ (M4 １－ｙ M4' ｙ ) １＋ｂ O ３ The range variable is also called a search variable, or simply a variable.

[0098] Here, range variables include categorical variables, discrete variables, and continuous variables. Categorical variables indicate the types of elements that make up a compound. For example, categorical variables are variables M3, M3', M4, and M4'. Each of variables M3 and M3' can take one of four option data "La, Al, Ga, In". In this case, there are six possible combinations of option data for each of variables M3 and M3' ( ４ C ２ = 6). Each of the variables M4 and M4' can take one of three option data "Ti, Zr, Hf". In this case, there are three combinations of option data that can be taken for each of the variables M4 and M4' ( ３ C ２= 3). Note that Li and O are pre-fixed elements and do not fall under the category of categorical variables.

[0099] Discrete variables indicate values ​​used to determine the composition ratio of a compound. The values ​​of multiple option data that can be taken for a discrete variable exist discretely from each other. The number of option data for a discrete variable is also smaller than that for a continuous variable. For example, discrete variables are variables a and b that determine the coefficients of elements. Variable a can take one of five option data: "0.0, 0.05, 0.1, 0.15, 0.2". There are five possible combinations of option data for variable a ( ５ C １ = 5). Variable b can take one of four option data "0.0, 0.1, 0.2, 0.3". There are four combinations of option data that can be taken for variable b ( ４ C １ = 4).

[0100] A continuous variable indicates a numerical value for determining the composition ratio of a compound. The number of option data for a continuous variable is greater than, for example, a discrete variable or a categorical variable, and the variable is suitable for detailed analysis of changes in characteristic values ​​in response to changes in option data. It is desirable that the number of option data for a continuous variable is five or more. For example, the continuous variables are variables x and y for determining the coefficients of elements. Each of the variables x and y can take one of eleven option data values, ranging from a minimum value of 0.0 to a maximum value of 1.0 with a step width of 0.1. In other words, there are eleven possible option data values ​​for each of the variables x and y ( １１ C １ =11).

[0101] In the example shown in FIG. 2 , the 11 choice data for the continuous variables are expressed using a minimum value of 0.0, a maximum value of 1.0, and a step size of 0.1. However, they may also be expressed using 11 coefficients, such as "0.0, 0.1, 0.2, 0.3, ..., 1.0." Furthermore, continuous variables may be set with a step size smaller than 0.1 so that search data, such as predicted characteristic values, change more continuously. Examples of smaller step sizes include 0.01 and 0.001. Continuous variables may be specified by the user. For example, the user may specify that each of the search variables x and y is a continuous variable. In this case, the user may specify the minimum and maximum values ​​of the search variables x and y, and values ​​previously stored in the characteristic display device 130 may be used as the step sizes for the search variables x and y.

[0102] In the formula of data representation shown in Figure 2, the ratios of M3 and M3', represented by the continuous variable x, have a negative correlation such that changing one ratio changes the other ratio. Similarly, the ratios of M4 and M4', represented by the continuous variable y, have a negative correlation such that changing one ratio changes the other ratio.

[0103] In other words, the variables M3, M3', M4, M4', a, b, x, and y are expressed as follows using option data:

[0104] M3∈{La, Al, Ga, In} M3'∈{La, Al, Ga, In} M4∈{Ti, Zr, Hf} M4'∈{Ti, Zr, Hf} a∈{0.0, 0.05, 0.1, 0.15, 0.2} b∈{0.0, 0.1, 0.2, 0.3} x∈{0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0} y∈{0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0}

[0105] Then, for each of the variables M3, M3', M4, M4', a, b, x, and y, one option data is selected from multiple option data corresponding to that variable, resulting in a combination of option data. The combinations are (M3, M3', M4, M4', a, b, x, y) = (La, Al, Ti, Zr, 0.1, 0.1, 0.1, 0.1), (La, Al, Ti, Zr, 0.1, 0.1, 0.1, 0.2), (La, Al, Ti, Zr, 0.1, 0.1, 0.1, 0.3), etc. In other words, there are many combinations. The composition formula, which is the structure of the compound, is uniquely expressed by such combinations of option data. Furthermore, the above-mentioned search range is a range that includes all of these combinations, and the above-mentioned search point can be said to represent one combination or the structure of a compound. A combination of option data that can be taken by each of a plurality of variables is also referred to as a combination of a plurality of variables. For example, a combination of option data that can be taken by each of variables x and y is also referred to as a combination of variables x and y.

[0106] By varying the values ​​or elements specified for these range variables to generate search points, different search points are generated, the collection of which represents the search range.

[0107] [Predicted Value Acquisition Unit 132] The predicted value acquisition unit 132 acquires a search range signal from the search range acquisition unit 131. The predicted value acquisition unit 132 then generates all possible combinations of option data for each of the multiple variables included in the search range indicated by the search range signal. The compositional formula of the compound is expressed by these combinations. Therefore, multiple compositional formulas or configurations are generated according to these combinations.

[0108] 3 is a diagram showing an example of a composition formula expressed by a combination of option data. Fig. 3 shows multiple composition formulas generated and the option data for each of the variables M3, M3', M4, M4', a, b, x, and y used in each of the multiple composition formulas. In the present disclosure, the coefficients of each element included in the composition formula (i.e., numerical values ​​indicating the composition ratios) may be indicated by subscripts or by regular characters, similar to the characters indicating the elements.

[0109] The data representation of the composition formula may indicate the order of the option data that can be taken for the variable, rather than the option data itself as shown in Figure 3. For example, if the five option data for the discrete variable a are "0.0, 0.05, 0.1, 0.15, 0.2," the option data is identified in that order. In this case, for example, instead of the discrete variable a = 0.1, the order "3" of the option data "0.1" for the discrete variable a may be used in the data representation of the composition formula.

[0110] The predicted value acquisition unit 132 acquires a predictor from the predictor database 120 and inputs the generated composition formulas into the predictor to acquire predicted property values ​​of compounds having those composition formulas.

[0111] FIG. 4 is a diagram showing an example of the obtained predicted property values ​​of each compound.

[0112] 4 (the right end of the figure), for each of the multiple compositional formulas generated from the search range, the predicted value acquisition unit 132 acquires a predicted property value of the compound having that compositional formula. For each of the multiple compositional formulas, the predicted value acquisition unit 132 associates the compositional formula with the predicted property value acquired for that compositional formula and outputs the associated information to the image processing unit 134. Note that the output compositional formula may be a combination of the multiple variables described above.

[0113] [Display Method Acquisition Unit 133] The display method acquisition unit 133 acquires an input signal indicating a display method from the input unit 110 as first display method information. Then, the display method acquisition unit 133 outputs the first display method information to the image processing unit 134. The display method indicated by the first display method information is a method of displaying predicted property values ​​of compounds represented by each composition formula.

[0114] [Image Processing Unit 134] Image processing unit 134 acquires, for each of the plurality of compounds, the composition formula and predicted property values ​​of that compound from predicted value acquisition unit 132. Furthermore, image processing unit 134 acquires first display method information from display method acquisition unit 133. Image processing unit 134 generates a map showing the predicted property values ​​of the plurality of compounds in accordance with the display method indicated by the first display method information, and displays an image including the map on display unit 140.

[0115] [Specific Examples of Maps] FIGS. 5A to 5C are diagrams showing examples of maps.

[0116] The image processing unit 134 generates, for example, a map Ma shown in FIG. 5A . This map Ma has coordinate axes A1 and A2. For example, a continuous variable x is assigned to the coordinate axis A1, and for example, a continuous variable y is assigned to the coordinate axis A2. Cells at each position indicated by the respective values ​​of the variables x and y in the map Ma are colored with a shading level corresponding to the predicted property value of a compound whose composition formula includes the respective values ​​of the variables x and y. The predicted property value is, for example, a band gap, and the map Ma is provided with a scale indicating the relationship between the predicted property value and the shading level.

[0117] For example, the display method acquisition unit 133 acquires first display method information in response to a user's input operation on the input unit 110. This first display method information indicates, for example, a variable x assigned to coordinate axis A1 and a variable y assigned to coordinate axis A2. Furthermore, the first display method information indicates, as designated option data, each of the option data for variables M3, M3', M4, M4', a, and b that are not assigned to a coordinate axis. Then, the display method acquisition unit 133 outputs the first display method information to the image processing unit 134.

[0118] The image processing unit 134 assigns the variable x to the coordinate axis A1 of the map Ma and the variable y to the coordinate axis A2 of the map Ma in accordance with the first display method information. Furthermore, the image processing unit 134 identifies the predicted property values ​​of each compound for which the variables M3, M3', M4, M4', a, and b each indicate designated choice data from the predicted property values ​​of the multiple compounds acquired by the predicted value acquisition unit 132. The image processing unit 134 then applies a color with a shade corresponding to the predicted property value of each identified compound to the position on the map Ma indicated by the respective values ​​of the variables x and y included in the composition formula of that compound.

[0119] For example, the color may be black or white, and the larger the predicted characteristic value, the darker the black color assigned to that predicted characteristic value on the map Ma. Conversely, the smaller the predicted characteristic value, the lighter the black color assigned to that predicted characteristic value on the map Ma. However, the color may be a color other than black or white. Furthermore, the image processing unit 134 may display the predicted characteristic values ​​on the map Ma using different colors instead of different shades of color. For example, the image processing unit 134 may assign red to the first largest predicted characteristic value, orange to the second largest predicted characteristic value, yellow to the third largest predicted characteristic value, white to the fourth largest predicted characteristic value, yellow-green to the fifth largest predicted characteristic value, light blue to the sixth largest predicted characteristic value, and blue to the seventh largest predicted characteristic value. The image processing unit 134 may then assign colors to the map Ma that correspond to the magnitudes of the predicted characteristic values ​​acquired from the predicted value acquisition unit 132. In this case, the map Ma is displayed like a heat map.

[0120] Furthermore, for the multiple compounds having predicted property values ​​represented by this map Ma, variables other than variables x and y, i.e., variables M3, M3', M4, M4', a, and b that are not used on the coordinate axes of map Ma, each indicate the same option data. That is, the option data for each of variables M3, M3', M4, M4', a, and b is fixed to designated option data, and predicted property values ​​for the compounds corresponding to the respective values ​​of variables x and y are displayed on map Ma. In other words, in map Ma, the combinations of option data for each of variables M3, M3', M4, M4', a, and b are fixed. For example, the option data for each of variables M3, M3', M4, M4', a, and b is fixed to designated option data, such as variables (M3, M3', M4, M4', a, b) = (La, Al, Ti, Zr, 0.05, 0.1).

[0121] Note that variables used on the coordinate axes of map Ma, such as variables x and y, may be called variable variables or visualized variables, and variables not used on the coordinate axes of map Ma, such as variables M3 and M4, may be called fixed variables, non-visualized variables, surplus variables, or non-utilized variables, etc. Furthermore, map Ma in this embodiment is also called image element map Ma.

[0122] 5B , the image processing unit 134 may interpolate the predicted characteristic values ​​and display a map Ma indicating the interpolated predicted characteristic values ​​on the display unit 140. That is, the image processing unit 134 calculates predicted characteristic values ​​between 121 (=11×11) points based on predicted characteristic values ​​corresponding to the 11 numerical values ​​of the variable x (0.0, 0.1, 0.2, ..., 10.0) and the 11 numerical values ​​of the variable y (0.0, 0.1, 0.2, ..., 10.0). The image processing unit 134 may interpolate the predicted characteristic values ​​using, for example, linear interpolation, bilinear interpolation, or bicubic interpolation.

[0123] 5A and 5B, the numerical range of the continuous variable x assigned to the coordinate axis A1 is equal to the numerical range of the continuous variable y assigned to the coordinate axis A2. In other words, the numerical range of the continuous variables x and y ranges from a minimum value of 0.0 to a maximum value of 1.0. Therefore, the shape of the map Ma shown in FIGS. 5A and 5B is square. However, the shape of the map Ma may be other shapes depending on the method of assigning the coordinate axes of the map Ma.

[0124] For example, the shape of the map Ma may be rectangular, or may be triangular as shown in FIG. 5C.

[0125] In the example of FIG. 5C, IN(N １－ｘ－ｙ Se ｘ As ｙ 5A and 5B , a triangular map Ma is displayed, which indicates the predicted property values ​​of a compound having a composition formula of the formula (x, y). As in the examples of FIGS. 5A and 5B , the cells at the positions indicated by the values ​​of the variables x and y in the map Ma are colored with a shade of color corresponding to the predicted property values ​​of a compound whose composition formula includes the values ​​of the variables x and y.

[0126] As described above, the property display device 130 in this embodiment includes a predicted value acquisition unit 132, a display method acquisition unit 133, and an image processing unit 134. The predicted value acquisition unit 132 acquires predicted property values ​​for each of a plurality of compounds. The display method acquisition unit 133 acquires first display method information indicating a display method for the predicted property values. The image processing unit 134 generates a map Ma indicating the predicted property values ​​for each of the plurality of compounds according to the first display method information. Furthermore, the image processing unit 134 generates and outputs an image including the map Ma. Here, the map Ma has coordinate axes indicating at least two variables, each of which is one of a plurality of variables used to represent the structure of the compound. For example, the image described above is output to and displayed on the display unit 140.

[0127] As a result, first display method information is acquired, and a map Ma showing predicted property values ​​for each of the multiple compounds is generated according to the display method indicated by the first display method information. As a result, if the display method indicated by the first display method information is set in accordance with the purpose of searching for new materials, displaying an image including the generated map Ma can appropriately support the search for new materials, i.e., material development. Furthermore, because the map Ma has the above-mentioned coordinate axes, the predicted property values ​​of each of the multiple compounds can be displayed at positions corresponding to the respective structures of the compounds. Furthermore, because the predicted property values ​​of each of the many compounds to be searched for are displayed as the map Ma, the predicted property values ​​can be appropriately visualized, allowing the user to easily recognize the relationship between the compound structures and the predicted property values. In other words, the user can easily grasp the overall picture of the predicted property values ​​in the map Ma. The search for new materials or new compounds is also referred to as material search.

[0128] The characteristic display device 130 does not necessarily have to include the display method acquisition unit 133. In this case, the image processing unit 134 outputs an image including a map Ma generated using the acquired predicted characteristic values. The map Ma has coordinate axes each showing at least two variables used to express the structure of a compound, and is a map showing the predicted characteristic values ​​of each of the multiple compounds. Even with this characteristic display device 130, the same effects as those described above can be obtained.

[0129] Furthermore, the first display method information may indicate at least one of a method for determining a color for indicating a predicted property value and a method for determining a display format for indicating the predicted property value as a display method for the predicted property value. For example, a color corresponding to the predicted property value is determined by the color determination method, and a color shading corresponding to the predicted property value is determined by the display format determination method. As a result, the image processing unit 134 in this embodiment generates, for each of a plurality of compounds, a map Ma that indicates a color or a color shading corresponding to the predicted property value of that compound.

[0130] This allows the user to easily grasp the magnitude of the predicted property value, for example, the difference in the predicted property values ​​between compounds with similar structures, because the predicted property value of each compound is displayed in a color or shade of color corresponding to the predicted property value.

[0131] The property display device 130 in this embodiment also includes a search range acquisition unit 131. The search range acquisition unit 131 acquires multiple variables used to express the structure of a compound and multiple option data items indicating possible values ​​or elements for each of the multiple variables. The predicted value acquisition unit 132 then acquires, for each combination of option data obtained by selecting one option data item from the multiple option data items for each of the multiple variables, a predicted property value for a compound having a structure corresponding to that combination. In other words, the search range acquisition unit 131 acquires, for each of the combinations described above, a predicted property value for a compound having a structure corresponding to that combination using a predetermined algorithm. For example, the predetermined algorithm is a predictor stored in the predictor database 120.

[0132] This allows for appropriate predicted property values ​​to be obtained. That is, the property values ​​of multiple compounds with different compositions can be appropriately predicted. Furthermore, when the predetermined algorithm is updated based on experiments on compounds that are conducted from time to time, the prediction accuracy of the property values ​​can be improved.

[0133] FIG. 6 is a diagram showing another example of the map Ma.

[0134] 6, the image processing unit 134 may discretize the predicted characteristic values ​​and apply colors of gradation corresponding to the discretized predicted characteristic values ​​to the map Ma. For example, the image processing unit 134 discretizes the predicted characteristic values ​​"0.5 to 4.0" at intervals of 0.25. Therefore, if the predicted characteristic value acquired from the predicted value acquisition unit 132 is, for example, 0.7, the image processing unit 134 replaces the predicted characteristic value with, for example, 0.75.

[0135] For example, the display method acquisition unit 133 acquires an input signal indicating the discretization of the predicted property values ​​and the intervals therebetween as first display method information from the input unit 110, and outputs the first display method information to the image processing unit 134. The image processing unit 134 discretizes the predicted property values ​​of each compound acquired from the predicted value acquisition unit 132 in accordance with the first display method information.

[0136] In this way, the image processing unit 134 in this embodiment discretizes the predicted characteristic values ​​and generates the map Ma indicating the discretized predicted characteristic values.

[0137] That is, in this embodiment, it is not limited to applying colors with gradations corresponding to the acquired predicted characteristic values ​​themselves to the map Ma, but it is also possible to apply colors with gradations corresponding to the discretized predicted characteristic values ​​to the map Ma as shown in Fig. 6. This improves the visibility of areas having similar predicted characteristic values ​​in the map Ma, makes it easier to grasp the change trend or distribution of the predicted characteristic values, and reduces the number of steps required to grasp the predicted characteristic values.

[0138] 7A to 7D are diagrams showing other examples of the map Ma.

[0139] 7A , the closer the predicted characteristic value is to a reference value, the lower the shading level the image processing unit 134 may set for that predicted characteristic value, and the farther the predicted characteristic value is from the reference value, the higher the shading level the image processing unit 134 may set for that predicted characteristic value. For example, the display method acquisition unit 133 acquires first display method information indicating the reference value from the input unit 110 and outputs the first display method information to the image processing unit 134. The image processing unit 134 sets the shading level for the predicted characteristic value in accordance with the reference value indicated in the first display method information.

[0140] In the example of FIG. 7A , the reference value is a value within a range of 2.65 to 2.85 eV. Note that a low shading level indicates a light color, and a high shading level indicates a dark color. Furthermore, the image processing unit 134 may assign the same color scheme to predicted characteristic values ​​when the predicted characteristic values ​​are greater than the reference value and when they are smaller than the reference value. For example, as shown in the example of FIG. 7A , predicted characteristic values ​​exceeding 2.85 eV are greater than the reference value, and predicted characteristic values ​​less than 2.65 eV are smaller than the reference value. Furthermore, in the map Ma, a region r1 showing predicted characteristic values ​​greater than 2.85 eV and a region r2 showing predicted characteristic values ​​less than 2.65 eV may be assigned the same color scheme. Conversely, the image processing unit 134 may assign different colors to predicted characteristic values ​​when the predicted characteristic values ​​are greater than the reference value and when they are smaller than the reference value. For example, in the map Ma, a region r1 where a predicted characteristic value exceeds 2.85 [ev] may be displayed in a reddish color, and a region r2 where a predicted characteristic value is less than 2.65 [ev] may be displayed in a bluish color. Furthermore, the image processing unit 134 sets a predetermined white color to a region where a predicted characteristic value equal to the reference value is displayed. However, the color set to the region is not limited to white, and may be another color such as red. Furthermore, the predicted characteristic value may be discretized as in the example of FIG. 6 , or may not be discretized. When the predicted characteristic value is discretized, the multiple regions in the map Ma are displayed as if separated by contour lines, for example.

[0141] As shown in FIG. 7B , the image processing unit 134 may apply a predetermined hatching pattern to a region r2 in the map Ma where the predicted characteristic value is smaller than the reference value. The pattern may be a striped pattern, or, as in the example of FIG. 7B , a pattern consisting of multiple dots. That is, the image processing unit 134 may superimpose multiple dots on the region r2. Alternatively, the image processing unit 134 may superimpose only multiple dots on the region r2 instead of a color with a shade corresponding to the predicted characteristic value.

[0142] The image processing unit 134 may superimpose a contour line L1 corresponding to the reference value on the map Ma, as shown in FIG. 7C . For example, as shown in FIG. 7C , the contour line L1 is superimposed on the boundary between a region r3 where the predicted characteristic value is within the reference value range and a region r2 where the predicted characteristic value is below that range. Note that the image processing unit 134 may superimpose the contour line L1 on the map Ma without superimposing multiple dots on the map Ma, or may superimpose only the contour line L1 on the map Ma. This allows the user to easily grasp the region where the predicted characteristic value is near the reference value or the region where the predicted characteristic value exceeds the reference value.

[0143] As shown in FIG. 7D , the image processing unit 134 may apply a predetermined color or a color of a predetermined shade to a region r4 (a striped region in the example of FIG. 7D ) in the map Ma where a predicted characteristic value that does not satisfy the condition is indicated. The color may be black. Alternatively, the image processing unit 134 may apply a predetermined hatching pattern to the region r4. In other words, the image processing unit 134 may superimpose a striped pattern on the region r2. The above-mentioned condition is, for example, that the predicted characteristic value is equal to or greater than a lower limit value. In the example of FIG. 7D , the lower limit value is 2.45 eV.

[0144] For example, the display method acquisition unit 133 acquires first display method information indicating the condition or lower limit value from the input unit 110 and outputs the first display method information to the image processing unit 134. The image processing unit 134 generates a map Ma, such as that shown in FIG. 7D , in accordance with the first display method information. Areas below the lower limit value are areas that do not need to be considered in material search, for example. Therefore, in the example of FIG. 7D , the user can easily identify areas that do not need to be considered in material search.

[0145] As described above, the first display method information in this embodiment indicates at least one of the following (a) to (e) as a display method for predicted characteristic values. (a) is a method of displaying predicted characteristic values ​​equal to the reference value on the map Ma using a predetermined color or shade of color. (b) is a method of displaying predicted characteristic values ​​greater than the reference value using a first series of colors and predicted characteristic values ​​smaller than the reference value using a second series of colors. (c) is a method of superimposing a boundary line on the boundary between a first region on the map Ma showing predicted characteristic values ​​equal to the reference value and a second region on the map Ma showing predicted characteristic values ​​different from the reference value. (d) is a method of displaying one of a third region on the map Ma showing predicted characteristic values ​​that satisfy a predetermined condition or a fourth region on the map Ma showing predicted characteristic values ​​that do not satisfy the predetermined condition using a predetermined color or shade of color. (e) is a method of superimposing a pattern of multiple dots or stripes on one of the third and fourth regions. For example, methods (a) and (b) are the display methods shown in Figures 7A to 7C, method (c) is the display method shown in Figure 7C, method (d) is the display method shown in Figure 7D, and method (e) is the method shown in Figures 7B to 7D. Note that the boundary line in (c) corresponds to the contour line L1 in Figure 7C.

[0146] This allows the user to easily grasp the areas where the characteristic prediction value is equal to the reference value, the areas where the characteristic prediction value is greater than or less than the reference value, the boundaries between these areas, or the areas where the specified conditions are met or not met.

[0147] In the above example, the reference value is indicated by the first display method information. That is, the reference value is a value specified by the user through an input operation on the input unit 110. However, the reference value may also be a value determined from the predicted characteristic values. For example, the reference value may be the average or median of all the predicted characteristic values ​​indicated on the map Ma.

[0148] That is, the reference value in this embodiment is the average or median of the predicted property values ​​of each of a plurality of compounds, or a value designated by the user.

[0149] This makes it possible to easily set appropriate reference values ​​for dividing the entire map Ma into a plurality of regions or for displaying the boundaries between these regions.

[0150] FIG. 8 is a diagram showing another example of the map Ma.

[0151] The image processing unit 134 may superimpose an arrow indicating the gradient of the predicted characteristic value on the map Ma, as shown in FIG. 8 . For example, the display method acquisition unit 133 acquires first display method information indicating the superimposition of the arrow from the input unit 110 and outputs the first display method information to the image processing unit 134. The image processing unit 134 calculates the gradient of the predicted characteristic value at each position on the map Ma in accordance with the first display method information and superimposes an arrow indicating the gradient at that position. The arrow indicates the direction of the gradient and has a length proportional to the magnitude of the gradient. This makes it easy to grasp the magnitude and direction of change in the predicted characteristic value according to the respective values ​​of the variables x and y.

[0152] That is, the image processing unit 134 in this embodiment identifies the gradient of the predicted characteristic value shown on the map Ma and superimposes an arrow indicating the direction and magnitude of the gradient on the map Ma. This allows the user to easily grasp the change in the predicted characteristic value relative to the variables x and y by looking at the arrow on the map Ma.

[0153] 9A and 9B are diagrams showing other examples of the map Ma.

[0154] As shown in FIG. 9A , the image processing unit 134 may display the best predicted property value among the predicted property values ​​of multiple compounds having the same combination of the values ​​of the variables x and y at a position on the map Ma corresponding to that combination. The best predicted property value is, for example, the maximum predicted property value. The best predicted property value is also called a best value, and may be the minimum predicted property value or the predicted property value closest to a predetermined reference value.

[0155] Specifically, if the identical combination of the values ​​of the variables x and y is the variable (x, y) = (0.1, 0.2), the image processing unit 134 identifies predicted property values ​​for multiple compounds each containing the variable (x, y) = (0.1, 0.2) in their composition formula. That is, the image processing unit 134 narrows down the predicted property values ​​for multiple compounds each containing the variable (x, y) = (0.1, 0.2) in their composition formula from among the predicted property values ​​for all compounds acquired by the predicted value acquisition unit 132. Next, the image processing unit 134 determines the best predicted property value from among the identified multiple predicted property values ​​as the best value. Then, the image processing unit 134 displays the determined best value at a position on the map Ma corresponding to the variable (x, y) = (0.1, 0.2) using a color with a shade corresponding to the best value. This determination of the best value is performed for each position on the map Ma, and the determined best value is displayed at that position on the map Ma.

[0156] Furthermore, the image processing unit 134 may identify predicted property values ​​for multiple compounds whose composition formulas include not only the variables (x, y) = (0.1, 0.2) but also the variables (M3, M3', M4, M4') = (La, Al, Ti, Zr). That is, the option data for the variables (M3, M3', M4, M4') included in the composition formulas of the multiple compounds may be the same or fixed. In this case, multiple composition formulas are expressed by using different combinations of the values ​​of the variables (a, b), and the best predicted property value is determined as the best value from the predicted property values ​​of the compounds represented by each of the multiple composition formulas.

[0157] For example, as shown in FIG. 9A , the best value (i.e., band gap) at each position on the map Ma exceeds 3.0 eV. In this case, the user can understand that even if the values ​​of the variables x and y are within the range of 0.0 to 1.0, by adjusting the values ​​of the variables a and b, a predicted characteristic value exceeding 3.0 eV can be obtained. Therefore, the predicted characteristic value can be displayed by focusing on the relationship between the important variables (e.g., variables x and y) used on the coordinate axes and the predicted characteristic value, without considering the variables (e.g., variables a and b) whose values ​​are not particularly important.

[0158] Furthermore, as shown in FIG. 9B , the image processing unit 134 may display the average value of the predicted property values ​​of a plurality of compounds having the same combination of the values ​​of the variables x and y at a position on the map Ma corresponding to that combination.

[0159] Specifically, when the identical combination of the numerical values ​​of the variables x and y is the variable (x, y) = (0.1, 0.2), the image processing unit 134 identifies predicted property values ​​for multiple compounds each containing the variable (x, y) = (0.1, 0.2) in their composition formula. That is, the image processing unit 134 narrows down the predicted property values ​​for multiple compounds each containing the variable (x, y) = (0.1, 0.2) in their composition formula from among the predicted property values ​​for all compounds acquired by the predicted value acquisition unit 132. Next, the image processing unit 134 calculates the average value of the identified multiple predicted property values. Then, the image processing unit 134 displays the calculated average value at a position on the map Ma corresponding to the variable (x, y) = (0.1, 0.2) using a color with a shade corresponding to the average value. This determination of the average value is performed for each position on the map Ma, and the calculated average value is displayed at that position on the map Ma.

[0160] Furthermore, the image processing unit 134 may specify predicted property values ​​for a plurality of compounds whose composition formulas include not only the variables (x, y) = (0.1, 0.2) but also the variables (M3, M3', M4, M4') = (La, Al, Ti, Zr). That is, the option data for the variables (M3, M3', M4, M4') included in the composition formulas of the plurality of compounds may be the same or fixed. In this case, a plurality of composition formulas are expressed by varying the combinations of the values ​​of the variables (a, b), and an average value of the predicted property values ​​for the compounds represented by each of the plurality of composition formulas is calculated.

[0161] Furthermore, a numerical range may be set for each of the variables a and b. For example, the display method acquisition unit 133 acquires first display method information indicating the variables (a, b) = (0.05, 0.1) and outputs the first display method information to the image processing unit 134. The image processing unit 134 sets the numerical ranges for each of the variables a and b according to the first display method information, such as variable a ∈ {0.0, 0.05, 0.1} and variable b ∈ {0.0, 0.1, 0.2}. That is, for each value of the variables a and b indicated by the first display method information, the image processing unit 134 includes that value and one value before and after that value in the numerical range. In this case, multiple composition formulas are expressed by varying the combinations of values ​​within the respective numerical ranges of the variables a and b, and the average value of the predicted property values ​​of the compounds represented by each of the multiple composition formulas is calculated. Note that the image processing unit 134 may calculate a differential value, variance, or the like instead of the average value. This makes it possible to display the predicted characteristic values ​​taking into consideration fluctuations in the predicted characteristic values ​​when the variables a and b are changed, and to grasp the robustness of the predicted characteristic values.

[0162] In this embodiment, if there is a non-utilized variable a other than the two variables x and y used on the coordinate axes A1 and A2 of the map Ma among the multiple variables, the first display method information indicates that the predicted property value is to be displayed using the non-utilized variable a as the display method for the predicted property value. This allows the predicted property value to be displayed not only using the two variables x and y used on the coordinate axes A1 and A2 of the map Ma, but also using the non-utilized variable a, thereby appropriately supporting materials development.

[0163] Specifically, the first display method information in this embodiment indicates that, when there is a non-utilized variable a other than the two variables x and y used for the coordinate axes A1 and A2 of the map Ma among the multiple variables, the non-utilized variable a is assigned a first value, a second value, or a value within a predetermined range of values ​​as a display method for the predicted property values. Here, (a) when a first value is assigned to the non-utilized variable a, the image processing unit 134 generates a map Ma showing the predicted property values ​​of each of the multiple compounds having a configuration represented by the non-utilized variable a representing the first value specified by the user. Also, (b) when a second value is assigned to the non-utilized variable a, the image processing unit 134 determines a second value such that the predicted property values ​​shown in the map Ma satisfy a predetermined condition, and generates a map Ma showing the predicted property values ​​of each of the multiple compounds having a configuration represented by the non-utilized variable a representing the determined second value. Furthermore, (c) when each numerical value within a predetermined numerical range is assigned to the non-utilized variable a, the image processing unit 134 calculates, for each position on the map Ma, the average of the predicted property values ​​of a plurality of compounds having a configuration expressed using two variables x and y, each of which indicates a numerical value corresponding to that position. The image processing unit 134 then generates a map Ma showing the average of the predicted property values ​​calculated for each position on the map Ma. Here, the non-utilized variables a of the plurality of compounds for which the average predicted property values ​​are calculated indicate different numerical values ​​within a predetermined numerical range. Non-utilized variables are variables that are not utilized on the coordinate axes of the map Ma, such as the discrete variable a, and are also referred to as non-visualized variables. Furthermore, when not only variable a but also variable b is present as a non-utilized variable, a first value, a second value, or a numerical value within a predetermined numerical range is assigned to each of variables a and b. Furthermore, in the case of (a) above, the map Ma shown in FIG. 5A is displayed, for example. That is, in the case of (a) above, the option data of the non-utilized variable such as the discrete variable a is designated or fixed to the first value as the above-mentioned designated option data by the user's input operation to the input unit 110. Moreover, the predetermined condition in the case of (b) above is, for example, a condition under which the characteristic prediction value becomes the above-mentioned best value.

[0164] As a result, in the case of (a), the user specifies a first value for the non-utilized variable, allowing the user to arbitrarily select the configuration of the compound having the predicted property value to be displayed on the map Ma. Furthermore, in the case of (b), by setting predetermined conditions as the conditions required for material search, it is possible to easily display a map Ma that satisfies the conditions required for material search, for example, without the user having to specify the value of the non-utilized variable. As a result, the efficiency of material search can be improved. In other words, material development can be appropriately supported. Furthermore, in the case of (c), changes in the values ​​that the non-utilized variable can take can be reflected in the predicted property value on the map Ma, thereby improving the robustness of the predicted property value with respect to the non-utilized variable.

[0165] [Processing Operation] FIG. 10 is a flowchart showing the processing operation of the display system 100 according to this embodiment.

[0166] (Step S111) First, the search range acquisition unit 131 acquires a search range signal output from the input unit 110 in response to an input operation by a user to the input unit 110. This search range signal is an input signal indicating the search range of a compound, for example, as shown in FIG. 2 . In other words, the search range acquisition unit 131 acquires the search range. Then, the search range acquisition unit 131 outputs the search range signal to the predicted value acquisition unit 132.

[0167] (Step S112) The predicted value acquisition unit 132 acquires the search range signal from the search range acquisition unit 131 and combines the option data for each of the multiple variables according to the search range indicated by the search range signal. For example, all possible combinations within the search range are generated. The predicted value acquisition unit 132 then inputs the composition formulas of each of the multiple compounds, i.e., the combinations, into a predictor in the predictor database 120 to acquire predicted property values ​​for those compounds. The predicted value acquisition unit 132 associates the predicted property values ​​of those compounds with the composition formulas of those compounds and outputs them to the image processing unit 134.

[0168] In this embodiment, the predicted value acquisition unit 132 acquires a predictor from the predictor database 120 and acquires a predicted property value using the predictor. However, the predicted value acquisition unit 132 may acquire predicted property values ​​for each of a plurality of compounds from at least one predictor for predicting the property values ​​of compounds, which is stored in the predictor database 120. Even in this case, it is possible to acquire an appropriate predicted property value for each of the plurality of compounds.

[0169] (Step S113) The display method acquisition unit 133 acquires first display method information output from the input unit 110 in response to an input operation by a user to the input unit 110. This first display method information indicates a display method for the characteristic prediction value, for example, as shown in Figures 5A to 9B. Then, the display method acquisition unit 133 outputs the first display method information to the image processing unit 134.

[0170] (Step S114) Image processing unit 134 acquires, for each of the plurality of compounds, the predicted property values ​​and composition formula of that compound from predicted value acquisition unit 132, and acquires first display method information from display method acquisition unit 133. Then, image processing unit 134 generates map Ma indicating the predicted property values ​​of each of the plurality of compounds in accordance with the display method indicated by the first display method information, and outputs an image including map Ma to display unit 140.

[0171] (Step S115) The display unit 140 acquires the image from the image processing unit 134 and displays the image.

[0172] By executing the processes of steps S111 to S115, the predicted property values ​​of the compound, which is the material, are displayed, that is, the properties are displayed.

[0173] (First Modification of Embodiment 1A) In the above embodiment, the variables x and y are used as the coordinate axes of map Ma, but variables a and b are not used. In this modification, image processing unit 134 generates a map using not only variables x and y but also variables a and b as coordinate axes.

[0174] FIG. 11 is a diagram showing an example of a map in this modified example.

[0175] 11, the image processing unit 134 generates an image map Mb consisting of an array of multiple maps Ma and displays it on the display unit 140. In this case, the image processing unit 134 assigns two variables that were not assigned to the coordinate axes A1 and A2 of the map Ma to two coordinate axes A3 and A4 of the image map Mb. The variables assigned to the two coordinate axes A3 and A4 are, for example, discrete variables a and b.

[0176] The map Ma is also called an image element map Ma to distinguish it from the image map Mb. Furthermore, the image element map Ma and the image map Mb are also simply called maps.

[0177] The discrete variable a has five option data values: 0.0, 0.05, 0.1, 0.15, and 0.2, while the discrete variable b has four option data values: 0.0, 0.1, 0.2, and 0.3. Therefore, there are 5 x 4 = 20 image element maps Ma for each combination of categorical variables in Figure 11. In other words, the image map Mb consists of an array of 5 x 4 image element maps Ma.

[0178] In the above example, in order to make it easier for the user to visually understand, when comparing the continuous variables x and y with the discrete variables a and b, the continuous variables x and y which have a larger number of combinations are assigned to the coordinate axes of the image element map Ma, and the discrete variables a and b which have a smaller number of combinations are assigned to the coordinate axes of the array of the image element map Ma. The assignment of each variable may be performed arbitrarily by the user, or the image processing unit 134 may calculate the number of combinations and automatically assign each variable.

[0179] (Variation 2 of Embodiment 1A) The range variables in the above embodiment include categorical variables M3, M3', M4, and M4', discrete variables a and b, and continuous variables x and y, but may also include other variables. The range variables in this variation include not only the above-mentioned variables but also one or more process variables that indicate conditions used in a compound production process, for example.

[0180] FIG. 12 is a diagram showing an example of process variables in this modification.

[0181] The process variables are also called process conditions, and include, for example, a variable Pa indicating a compound calcination method, a variable Pb indicating a compound calcination time, and a variable Pc indicating a compound calcination temperature. The variable Pa indicates, for example, a solid-state reaction method or a ball mill as option data. The variable Pb indicates, for example, 1 hour, 2 hours, or 3 hours as option data. The variable Pc indicates, for example, 100°C, 110°C, 120°C, 130°C, 140°C, 150°C, or 160°C as option data.

[0182] For example, the search range acquisition unit 131 acquires each option data that can be taken by the variables M3, M3', M4, M4', a, b, x, and y, and each option data that can be taken by the variables Pa, Pb, and Pc. The predicted value acquisition unit 132 generates all combinations of option data that can be taken by M3, M3', M4, M4', a, b, x, y, Pa, Pb, and Pc. The predicted value acquisition unit 132 then acquires a predictor from the predictor database 120 and inputs all the generated combinations (i.e., composition formulas and process conditions) into the predictor to acquire predicted property values ​​for each of the multiple compounds. The composition formula of a compound, or the composition formula and process conditions of a compound, are hereinafter also referred to as the compound configuration. The process variables are not limited to the variables Pa, Pb, and Pc, and may be any variable. The image processing unit 134 may assign one or more of the variables Pa, Pb, and Pc to the coordinate axes of the map Ma or Mb.

[0183] This allows the search range to include not only the composition formula but also the process conditions, making it possible to obtain predicted property values ​​for compounds generated under conditions that match the process conditions in the search range, and to display a map showing the obtained predicted property values.

[0184] (Embodiment 1B) As in Embodiment 1A, the display system in this embodiment displays predicted property values ​​for each of a plurality of compounds in the form of a map, and also superimposes experimental property values ​​for each of one or more compounds on the map. Note that, among the components in this embodiment, the same components as those in Embodiment 1A are assigned the same reference numerals as those in Embodiment 1A, and detailed description thereof will be omitted.

[0185] [Configuration of Display System 200] Fig. 13 is a block diagram showing an example of the configuration of a display system 200 according to this embodiment. The display system 200 shown in Fig. 13 includes an input unit 110, a predictor database 120, a characteristic display device 230, a display unit 140, and an experiment database (DB) 150. The characteristic display device 230 is an example of an information display device.

[0186] The characteristic display device 230 in this embodiment acquires experimental characteristic values, which are characteristic values ​​of compounds obtained through experiments, from the experimental database 150, generates a map on which the experimental characteristic values ​​are superimposed, and displays the map on the display unit 140. Such a characteristic display device 230 includes a search range acquisition unit 131, a predicted value acquisition unit 132, a display method acquisition unit 133, an experimental value acquisition unit 232, and an image processing unit 234. The characteristic display device 230 may be composed of a processor such as a CPU and a memory. In this case, the processor functions as the characteristic display device 230 by, for example, executing a computer program stored in the memory. The memory may be volatile or nonvolatile, or may be composed of a volatile memory and a nonvolatile memory.

[0187] [Experimental Database 150] The experimental database 150 is a recording medium that stores experimental data indicating the composition formula (i.e., chemical formula) of a compound, compound identification information of the compound, and experimental characteristic values ​​of the compound. This recording medium may be, for example, a hard disk drive, a random access memory (RAM), a read-only memory (ROM), or a semiconductor memory. The recording medium may be either volatile or non-volatile. If the search range includes process variables, the experimental data may also include experimental process information indicating the process conditions used in the experiment.

[0188] FIG. 14 is a diagram showing an example of experimental data stored in the experimental database 150. As shown in the example of FIG. 14, the experimental data indicates, for each of a plurality of compounds, the composition formula of the compound, compound identification information, and experimental characteristic values. The compound identification information is a so-called ID, and may be any information that can identify the compound, such as a name, a symbol, or a number. For example, the experimental data may include the composition formula of the compound "Li １．４５ La ０．０４５ Ti １．１ Al ０．００５ O ３14 , the compound ID "000001-00001-001" and the compound's experimental characteristic value (exp.data) "2.349" are shown. Note that this composition formula is defined as (M3, M3', M4, M4', a, b, x, y) = (La, Al, Ti, Zr, 0.05, 0.1, 0.1, 0). The ID "000001-00001-001" in FIG. 14 is composed of three levels of numbers (000001, 00001, 001). For example, the IDs are registered so that first-level numbers "000001" to "000005" are assigned to five chemical formulas with different composition ranges for the elements Zr and Ti contained in the compound. Assigning IDs that allow classification of the composition formulas of compounds in this way facilitates management of the experiment database 150 by multiple users. 14 is selected arbitrarily depending on the compound. For example, the experimental characteristic value for a battery material is conductivity, and the experimental characteristic value for a thermoelectric conversion material is a thermoelectric conversion figure of merit, etc.

[0189] The experimental data may indicate process conditions and analytical information used to produce the compound. The process conditions include, for example, a calcination method, a calcination time, and a calcination temperature. The analytical information includes, for example, the crystal structure of the compound, the type of raw material, and whether or not any raw material remains. The presence or absence of any raw material remains is information indicating whether or not any raw material used for synthesizing or producing the compound remains after the synthesis or production of the compound.

[0190] [Experimental Value Acquisition Unit 232] The experimental value acquisition unit 232 acquires the composition and experimental characteristic values ​​of each of one or more compounds that have been tested from the experimental database 150. The composition may be the composition formula of the compound, or the composition formula and process conditions. For example, the experimental value acquisition unit 232 acquires a search range from the search range acquisition unit 131 and acquires the experimental characteristic values ​​of compounds having a composition included in the search range together with the composition. The experimental value acquisition unit 232 outputs the composition and experimental characteristic values ​​of each of the one or more compounds to the image processing unit 234.

[0191] [Display Method Acquisition Unit 133] The display method acquisition unit 133 acquires first display method information indicating a display method for predicted characteristic values ​​and second display method information indicating a display method for experimental characteristic values ​​from the input unit 110. Then, the display method acquisition unit 133 outputs the first display method information and the second display method information to the image processing unit 234.

[0192] [Image Processing Unit 234] The image processing unit 234 acquires the structures and predicted property values ​​of each of the multiple compounds from the predicted value acquisition unit 132, and acquires the structures and experimental property values ​​of each of one or more compounds from the experimental value acquisition unit 232. Furthermore, the image processing unit 234 receives first display method information and second display method information from the display method acquisition unit 133. The image processing unit 234 generates a map showing the predicted property values ​​of the multiple compounds according to the display method indicated by the first display method information. Furthermore, the image processing unit 234 superimposes the experimental property values ​​of each of the one or more compounds at positions on the map corresponding to the structures of the compounds. At this time, the image processing unit 234 superimposes the experimental property values ​​on the map according to the display method indicated by the second display method information. The image processing unit 234 then outputs an image including the map on which the experimental property values ​​of the one or more compounds are superimposed to the display unit 140. The map may be an image element map Ma or an image map Mb.

[0193] [Display Unit 140] The display unit 140 acquires an image from the image processing unit 234 and displays the image, that is, an image including a map on which experimental characteristic values ​​of one or more compounds are superimposed.

[0194] [Specific Example of Map] FIG. 15 is a diagram showing an example of a map.

[0195] The image processing unit 234 superimposes the characteristic experimental value on the image element map Ma of embodiment 1A as a mark such as a circle. For example, the color or color shading of the mark such as a circle indicates the characteristic experimental value. In other words, the image processing unit 234 superimposes a mark having a color or color shading corresponding to the characteristic experimental value of the compound at a position on the image element map Ma corresponding to the above-mentioned combination indicating the composition of the compound generated by the experiment. Note that the superimposition of the mark is also referred to as the superimposition of the characteristic experimental value. Furthermore, an example of the characteristic experimental value in the present disclosure is the band gap of the compound obtained by the experiment, but is not limited to this.

[0196] This makes it possible to highlight the difference between the color or color shading of the mark of the characteristic experimental value and the surrounding color or color shading when the characteristic experimental value deviates from the predicted characteristic value. As a result, it becomes easier to visually understand that the characteristic experimental value deviates from the predicted characteristic value. In other words, if the color or color shading of the mark of the characteristic experimental value and the color or color shading of the predicted characteristic value are visually the same, it becomes immediately visually understandable that the characteristic experimental value and the predicted characteristic value are substantially the same.

[0197] As described above, the characteristic display device 230 in this embodiment includes an experimental value acquisition unit 232 that acquires experimental characteristic values ​​of one or more compounds that have been tested. The image processing unit 234 then superimposes the experimental characteristic values ​​of one or more compounds that have been tested at positions on the image element map Ma that correspond to the configuration of the compound, and generates an image including the image element map Ma on which the experimental characteristic values ​​are superimposed.

[0198] This allows easy comparison of predicted property values ​​and experimental property values ​​within the range shown in the image element map Ma, thereby improving the efficiency of material search.

[0199] The characteristic display device 230 in this embodiment also includes a display method acquisition unit 133 that acquires second display method information indicating a display method for the characteristic experimental value. The image processing unit 234 then superimposes the characteristic experimental value on the image element map Ma in accordance with the second display method information.

[0200] As a result, the characteristic experimental values ​​are superimposed on the image element map Ma in accordance with the display method indicated by the second display method information, and therefore, the characteristic experimental values ​​can be displayed in a manner that suits the user's purpose for material exploration, depending on the setting of the display method, thereby improving the efficiency of material exploration and appropriately supporting material development.

[0201] FIG. 16 is a diagram showing another example of the map.

[0202] As shown in FIG. 16 , the image processing unit 234 may match (1) the correspondence between the experimental characteristic values ​​and the corresponding gradations, and (2) the correspondence between the predicted characteristic values ​​and the corresponding gradations. That is, the image processing unit 234 may match the gradation scale for the experimental characteristic values ​​with the gradation scale for the predicted characteristic values. As a result, when the predicted characteristic values ​​and the experimental characteristic values ​​are the same, the same gradation is assigned to those values, and the color of that gradation is applied to the image element map Ma. As a result, the degree of similarity between the predicted characteristic values ​​and the experimental characteristic values ​​can be easily grasped.

[0203] That is, in the characteristic display device 230 of this embodiment, predicted characteristic values ​​are displayed on the image element map Ma in a first display manner, which is a color or a shade of color corresponding to the predicted characteristic value. Furthermore, experimental characteristic values ​​are superimposed on the image element map Ma as marks having a second display manner, which is a color or a shade of color corresponding to the experimental characteristic value. In this case, the second display manner information indicates, as a display method for the experimental characteristic values, that the scale of the second display manner for the experimental characteristic values ​​is to be made to match the scale of the first display manner for the predicted characteristic values.

[0204] This allows the predicted characteristic values ​​and the experimental characteristic values ​​to be on the same scale, making it easier to compare the predicted characteristic values ​​with the experimental characteristic values.

[0205] 16, the image processing unit 234 may discretize the predicted characteristic values ​​while matching the grayscale scale for the experimental characteristic values ​​with the grayscale scale for the predicted characteristic values, but not the experimentally predicted values. The experimental characteristic values ​​are often acquired intensively in a specific region on the image element map Ma. Furthermore, these experimental characteristic values ​​are often similar. Therefore, the predicted characteristic values ​​are discretized to improve visibility, but the experimental characteristic values ​​are not discretized, thereby making it easy to read subtle differences in the experimental characteristic values.

[0206] FIG. 17 is a diagram showing another example of the map.

[0207] 17, when marks of a plurality of characteristic experimental values ​​overlap, the image processing unit 234 may arrange the mark indicating the characteristic experimental value closer to the front (i.e., the front side) as the characteristic experimental value is better. In other words, the image processing unit 234 may determine the order of overlapping of the characteristic experimental values.

[0208] For example, the display method acquisition unit 133 acquires second display method information indicating a method for arranging a plurality of marks as a method for displaying the characteristic experimental value from the input unit 110, and outputs the second display method information to the image processing unit 234. Upon acquiring the second display method information from the display method acquisition unit 133, the image processing unit 234 arranges the mark indicating the characteristic experimental value closer to the front (i.e., the front side) in accordance with the second display method information, the better the characteristic experimental value is. Note that a good characteristic experimental value may mean a large characteristic experimental value, a small characteristic experimental value, or close to the predicted characteristic value shown in the same position as the characteristic experimental value on the image element map Ma.

[0209] This makes it possible to present information useful for material search. In other words, when the characteristic experimental values ​​are superimposed on the image element map Ma, marks of the characteristic experimental values ​​are superimposed at the same position by the number of variables not used in the coordinate axes of the image element map Ma. Furthermore, multiple marks placed in positions close to each other on the image element map Ma are superimposed. When multiple marks are superimposed in this way, all or part of the marks other than the forefront mark are hidden by the forefront mark. In contrast, in the example shown in FIG. 17, the mark of the best characteristic experimental value is displayed at the forefront, making it possible to present information useful for material search.

[0210] In this manner, in the characteristic display device 230 of this embodiment, the second display method information indicates, as a characteristic experimental value display method, a rule that specifies the order of overlapping marks when multiple characteristic experimental values ​​are superimposed as overlapping marks on the image element map Ma. The image processing unit 234 then superimposes the marks of the multiple characteristic experimental values ​​on the image element map Ma in an overlapping manner according to the rule. The rule also specifies that (a) the larger the characteristic experimental value, the closer the mark for that characteristic experimental value is to be placed to the front side; (b) the closer the characteristic experimental value is to a predetermined value, the closer the mark for that characteristic experimental value is to be placed to the front side; or (c) the closer the characteristic experimental value is to the predicted characteristic value indicated at the position on the image element map Ma where the characteristic experimental value is superimposed, the closer the mark for that characteristic experimental value is to be placed to the front side.

[0211] This prevents marks indicating good experimental characteristic values, such as large experimental characteristic values, experimental characteristic values ​​close to a predetermined value, or experimental characteristic values ​​close to a predicted value, from being obscured by marks indicating other experimental characteristic values, thereby improving the efficiency of material search.

[0212] FIG. 18 is a diagram showing another example of the map.

[0213] As described above, the experimental value acquisition unit 232 acquires experimental characteristic values ​​of compounds having a structure included in the search range, but may acquire all experimental characteristic values ​​shown in the experimental data regardless of the search range. In this case, the structure of the compound corresponding to the experimental characteristic value acquired by the experimental value acquisition unit 232 may not match the structure on the image element map Ma. Specifically, as shown in FIG. 18 , the composition formula of the compound corresponding to the experimental characteristic value acquired by the experimental value acquisition unit 232 is the composition formula of a nitride, for example, Li １．４５ (La ０．５ Ga ０．５ ) ０．０５ (Ti ０．５ Zr ０．５ ) １．１ N ３ On the other hand, the composition formula shown in the image element map Ma is the composition formula of an oxide, for example, Li １．４５ (La １－ｘ Ga ｘ ) ０．０５ (Ti １－ｙ Zr ｙ ) １．１ O ３ If we focus on the variables x and y, the composition formula of the nitride is "Li １．４５ (La ０．５ Ga ０．５ ) ０．０５ (Ti ０．５ Zr ０．５ ) １．１ N ３ " corresponds to the position (x, y) = (0.5, 0.5) in the image element map Ma. However, the composition formula of the nitride does not match the composition formula of the oxide, and there is no position on the image element map Ma that corresponds to the composition formula of the nitride.

[0214] In such a case, the image processing unit 234 does not superimpose a mark indicating the experimental characteristic value corresponding to the composition formula of such a nitride on the image element map Ma.

[0215] Furthermore, when the search range and the experimental data each contain process conditions, the process conditions of the compound corresponding to the experimental characteristic value acquired by the experimental value acquisition unit 232 may not match the process conditions corresponding to the image element map Ma. The process conditions corresponding to the image element map Ma are one combination of process variables used to acquire all the predicted characteristic values ​​shown in the image element map Ma. In a specific example, the baking time of "5 hours" of the compound corresponding to the experimental characteristic value does not match the baking time of "10 hours" of the variable Pb used to acquire all the predicted characteristic values ​​shown in the image element map Ma.

[0216] In such a case, the image processing unit 234 does not superimpose a mark indicating the characteristic experimental value corresponding to the baking time of "5 hours" on the image element map Ma corresponding to the baking time of "10 hours."

[0217] In this way, by not superimposing characteristic experimental values ​​that do not correspond to the image element map Ma, it is possible to prevent misunderstandings that may arise from the superimposition of the characteristic experimental values.

[0218] Alternatively, the image processing unit 234 may set a new variable for the map to indicate the composition formula of the nitride, and thereby superimpose the experimental characteristic values ​​of the nitride on the map.

[0219] Alternatively, the image processing unit 234 may superimpose a characteristic experimental value that satisfies a predetermined condition on the image element map Ma, even if the composition formula and process conditions corresponding to the characteristic experimental value do not match the composition formula and process conditions corresponding to the image element map Ma. The predetermined condition may be, for example, a condition that the portion of the composition formula and process conditions corresponding to the characteristic experimental value that does not match the composition formula and process conditions corresponding to the image element map Ma is a portion designated by a user. For example, the element N in the composition formula of the above-mentioned nitride is different from the element O in the composition formula corresponding to the image element map Ma. Therefore, if the element O is a portion designated by the user, the image processing unit 234 may superimpose the characteristic experimental value corresponding to the composition formula of the nitride at a position (x, y) = (0.5, 0.5) on the image element map Ma. The position (x, y) = (0.5, 0.5) is the position on the image element map Ma that corresponds to the configuration closest to the nitride composition formula. Alternatively, the predetermined condition may be a condition that the composition formula and process conditions corresponding to the characteristic experimental value are within a search range.

[0220] Alternatively, the image processing unit 234 may calculate the similarity between the composition formula and process conditions corresponding to the experimental characteristic value and the composition formula and process conditions corresponding to the image element map Ma, and if the similarity is equal to or greater than a threshold, superimpose the experimental characteristic value on the image element map Ma. For example, the image processing unit 234 may represent the composition formula and process conditions corresponding to the experimental characteristic value as a vector, represent the composition formula and process conditions corresponding to the image element map Ma as a vector, and calculate the norm of the difference between these vectors as the distance. The similarity between the composition formula and process conditions corresponding to the experimental characteristic value and the composition formula and process conditions corresponding to the image element map Ma increases as the distance decreases, and decreases as the distance increases. If the distance is equal to or less than a threshold, i.e., if the similarity is equal to or greater than a threshold, the image processing unit 234 superimposes the experimental characteristic value on the image element map Ma. Note that the vector may include the coefficients of each element in the periodic table and the values ​​of multiple process variables as elements. In addition, the image processing unit 234 may normalize the distance so that the maximum value of the distance is 1 and the minimum value is 0, determine whether the normalized distance is below a threshold value, and if it is below the threshold value, superimpose the characteristic experimental value on the image element map Ma.

[0221] As described above, in the characteristic display device 230 of this embodiment, when there is no position on the image element map Ma corresponding to the structure of a compound having a characteristic experimental value, the second display method information indicates, as a display method for the characteristic experimental value, either (a) to superimpose the characteristic experimental value on the image element map Ma at a position corresponding to the structure closest to the structure of the compound having the characteristic experimental value, or (b) to not superimpose the characteristic experimental value on the image element map Ma. In this case, the image processing unit 234 performs processing related to superimposing the characteristic experimental value on the image element map Ma in accordance with the second display method information.

[0222] As a result, in the case of (a), even if the experimental characteristic value of a compound having a configuration that does not correspond to any position on the image element map Ma, the experimental characteristic value is superimposed at the position corresponding to the configuration closest to the configuration. Therefore, the experimental characteristic value of a compound having a configuration that does not correspond to the image element map Ma is also displayed in appropriate association with the image element map Ma, thereby improving the efficiency of material search by the user. Furthermore, in the case of (b), the experimental characteristic value that does not correspond to the image element map Ma is not superimposed, thereby reducing misunderstandings that may arise from the superimposition of the experimental characteristic value.

[0223] In the above example, the image processing unit 234 determines whether or not to superimpose the characteristic experimental value acquired by the experimental value acquisition unit 232. However, the experimental value acquisition unit 232 may acquire only the characteristic experimental value to be superimposed on the image element map Ma from all the characteristic experimental values ​​indicated in the experimental data in the experimental database 150 and output the acquired characteristic experimental value to the image element map Ma. In this case, the experimental value acquisition unit 232 determines whether or not the characteristic experimental value indicated in the experimental data is the characteristic experimental value to be superimposed on the image element map Ma, similar to the above-mentioned determination method by the image processing unit 234. Then, the image processing unit 234 superimposes the characteristic experimental value acquired by the experimental value acquisition unit 232 on the image element map Ma without determining whether or not to superimpose the characteristic experimental value.

[0224] FIG. 19 is a diagram showing another example of the map.

[0225] The image processing unit 234 may determine whether to superimpose the experimental characteristic value on the image element map Ma based on the experimental characteristic value itself. For example, the display method acquisition unit 133 acquires second display method information indicating an experimental threshold from the input unit 110 and outputs the second display method information to the image processing unit 234. The experimental threshold is, for example, 2.0 eV. The image processing unit 234 changes, for example, the image element map Ma shown in FIG. 19(a) to the image element map Ma shown in FIG. 19(b) in accordance with the second display method information. In the image element map Ma shown in FIG. 19(a), experimental characteristic values ​​in the band gap range of 0.0 to 0.35 eV are superimposed. On the other hand, in the image element map Ma shown in FIG. 19(b), experimental characteristic values ​​with a band gap of 2.0 eV or more are superimposed, and experimental characteristic values ​​with a band gap of less than 2.0 eV are not superimposed. That is, the image processing unit 234 determines that, among the experimental characteristic values ​​in the band gap range of 0.0 to 3.5 eV, experimental characteristic values ​​less than 2.0 eV should not be superimposed on the image element map Ma, and determines that experimental characteristic values ​​equal to or greater than 2.0 eV should be superimposed on the image element map Ma. This narrows down the experimental characteristic values. As a result, the image processing unit 234 generates the image element map Ma shown in FIG. 19(b) and displays an image including the image element map Ma on the display unit 140.

[0226] As described above, in the property display device 230 of the present embodiment, the second display method information indicates, as a display method of the property experimental values, that among the property experimental values ​​of one or more compounds that have been tested, property experimental values ​​that are equal to or greater than a predetermined first threshold value are superimposed on the image element map Ma, and property experimental values ​​that are less than the first threshold value are not superimposed on the image element map Ma. For example, the first threshold value is the above-mentioned experimental threshold value, and in a specific example, it is 2.0 [eV].

[0227] As a result, for example, when many experimental property values ​​are obtained, unimportant experimental property values ​​that would be treated as noise are not superimposed on the image element map Ma, but only important experimental property values ​​can be narrowed down from the many obtained experimental property values ​​and superimposed on the image element map Ma. As a result, the important experimental property values ​​superimposed on the image element map Ma can be made easier to see, and the efficiency of material search can be improved.

[0228] FIG. 20A is a diagram showing another example of the map.

[0229] As shown in FIG. 20A , the image processing unit 234 may emphasize and superimpose characteristic experimental values ​​that satisfy a predetermined condition. The predetermined condition is that the characteristic experimental values ​​have been obtained through experiments conducted within a certain period of time in the most recent period, i.e., experiments conducted within a certain period of time from the present time. The certain period is, for example, two weeks. Alternatively, the predetermined condition is that the characteristic experimental values ​​have been obtained through a predetermined number of experiments conducted most recently. The predetermined number of experiments most recently conducted is, for example, three.

[0230] That is, the display method acquisition unit 133 acquires second display method information indicating the above-mentioned predetermined conditions from the input unit 110 and outputs the second display method information to the image processing unit 234. In accordance with the second display method information, the image processing unit 234 superimposes characteristic experimental values ​​that satisfy the predetermined conditions on the image element map Ma in a more emphasized manner than characteristic experimental values ​​that do not satisfy the predetermined conditions, as shown in FIG. 20A . Specifically, characteristic experimental values ​​obtained from experiments conducted over the last two weeks, i.e., from the current time to two weeks ago, are superimposed on the image element map Ma as highly emphasized marks, and other characteristic experimental values ​​are superimposed on the image element map Ma as small marks. Alternatively, the three most recent characteristic experimental values, i.e., three characteristic experimental values ​​obtained from the three most recent experiments, are superimposed on the image element map Ma as highly emphasized marks. On the other hand, characteristic experimental values ​​obtained from experiments conducted three years prior to the current time are superimposed on the image element map Ma as small marks.

[0231] FIG. 20B is a diagram showing another example of the map.

[0232] The image processing unit 234 may emphasize and superimpose the experimental characteristic value that satisfies the condition of being larger than the reference value, as shown in Fig. 20B. In other words, in this case, the predetermined condition is that the experimental characteristic value is larger than the reference value.

[0233] For example, the display method acquisition unit 133 acquires second display method information indicating a reference value from the input unit 110 and outputs the second display method information to the image processing unit 234. In accordance with the second display method information, the image processing unit 234 superimposes experimental characteristic values ​​greater than the reference value on the image element map Ma, emphasizing them more than experimental characteristic values ​​equal to or less than the reference value, as shown in FIG. 20B . Specifically, experimental characteristic values ​​greater than the reference value are superimposed on the image element map Ma as highly emphasized marks, and experimental characteristic values ​​equal to or less than the reference value are superimposed on the image element map Ma as small marks. Note that, conversely to the above, the predetermined condition may be that the experimental characteristic value is smaller than the reference value. Furthermore, the reference value may be the reference value used in the examples of FIGS. 7A to 7D .

[0234] Alternatively, the image processing unit 234 may emphasize and superimpose experimental characteristic values ​​that satisfy the condition that the difference from the predicted characteristic values ​​is less than a difference threshold, as shown in Fig. 20B . That is, in this case, the predetermined condition is that the difference between the experimental characteristic value and the predicted characteristic value shown at the same position on the image element map Ma is less than a difference threshold. When the predicted characteristic value and the experimental characteristic value are band gaps, the difference threshold is, for example, 0.5 eV.

[0235] For example, the display method acquisition unit 133 acquires second display method information indicating the difference threshold from the input unit 110 and outputs the second display method information to the image processing unit 234. In accordance with the second display method information, the image processing unit 234 superimposes on the map Ma, as shown in FIG. 20B , experimental characteristic values ​​whose differences from the predicted characteristic values ​​are less than the difference threshold, more emphasizing them than experimental characteristic values ​​whose differences are equal to or greater than the difference threshold. Specifically, experimental characteristic values ​​whose differences from the predicted characteristic values ​​are less than the difference threshold are superimposed on the image element map Ma as highly emphasized marks, and experimental characteristic values ​​whose differences from the predicted characteristic values ​​are equal to or greater than the difference threshold are superimposed on the image element map Ma as small marks. Note that, conversely to the above, the predetermined condition may be that the difference between the experimental characteristic values ​​and the predicted characteristic values ​​displayed at the same position on the image element map Ma is equal to or greater than the difference threshold.

[0236] In this manner, in the property display device 230 of this embodiment, the second display method information indicates, as a display method of the property experimental values, that property experimental values ​​that satisfy predetermined conditions among the property experimental values ​​of one or more compounds that have been tested are superimposed on the image element map Ma in a manner that emphasizes the property experimental values ​​that do not satisfy the predetermined conditions. In this case, the image processing unit 234 superimposes the property experimental values ​​that satisfy the predetermined conditions on the image element map Ma in a manner that emphasizes the property experimental values.

[0237] This allows the user to visually and easily determine whether or not each characteristic experimental value displayed on the image element map Ma satisfies a predetermined condition, thereby improving the efficiency of material search.

[0238] Furthermore, in the characteristic display device 230 of this embodiment, the above-mentioned predetermined condition is (a) that the experimental characteristic value is a characteristic experimental value obtained within a predetermined period of time from the present, (b) that the experimental characteristic value is one of a predetermined number of most recently obtained experimental characteristic values, (c) that the experimental characteristic value is equal to or greater than a predetermined second threshold, or (d) that the difference between the experimental characteristic value and a predicted characteristic value obtained for a compound having the same structure as the compound having the experimental characteristic value is equal to or greater than a predetermined third threshold or less than the third threshold. For example, the second threshold is the above-mentioned reference value, and the third threshold is the above-mentioned difference threshold.

[0239] As a result, in (a) and (b), the user can visually and easily grasp new experimental characteristic values. In addition, in (c), the user can visually and easily grasp important experimental characteristic values, for example. In (d), the user can visually and easily grasp experimental characteristic values ​​that are close to predicted characteristic values.

[0240] FIG. 21 is a diagram showing another example of the map.

[0241] As shown in FIG. 21 , the image processing unit 234 may superimpose marks of experimental characteristic values ​​on the image element map Ma in a manner corresponding to the non-utilized variables corresponding to the experimental characteristic values. Non-utilized variables are associated with the experimental characteristic values ​​of each compound in the experimental data of the experimental database 150, for example. Non-utilized variables are, for example, variables indicating process conditions, such as the firing temperature, for example. Furthermore, non-utilized variables may be variables indicating a different type of characteristic from the experimental characteristic values. In other words, non-utilized variables are variables that are not used on the coordinate axes of the image element map Ma.

[0242] For example, the display method acquisition unit 133 acquires second display method information from the input unit 110, which prompts a display according to the non-utilized variables, and outputs the second display method information to the image processing unit 234. The experimental value acquisition unit 232 outputs the non-utilized variables of each compound, along with the structure and experimental characteristic values ​​of the compound, to the image processing unit 234. In accordance with the second display method information, the image processing unit 234 forms a mark for the experimental characteristic value in a manner corresponding to the data indicated by the non-utilized variable corresponding to the experimental characteristic value, and superimposes the mark on the image element map Ma, as shown in FIG. 21 . If the non-utilized variable is a firing temperature, the data indicated by the non-utilized variable is, for example, 500°C, 750°C, 1000°C, etc. Specifically, the image processing unit 234 forms a circular mark for the experimental characteristic values ​​of compounds fired at a firing temperature of 500°C, a triangular mark for the experimental characteristic values ​​of compounds fired at a firing temperature of 750°C, and a rectangular mark for the experimental characteristic values ​​of compounds fired at a firing temperature of 1000°C.

[0243] Although the above example shows one non-utilized variable, the number may be two or more. In this case, the image processing unit 234 calculates the product of the number of cases of two or more non-utilized variables and determines the number of different mark forms corresponding to the product. For example, the baking method, the baking temperature, and the baking time are three non-utilized variables. If the number of cases for the baking method is two, the number of cases for the baking time is seven, and the number of cases for the baking temperature is three, the image processing unit 234 determines 2 x 7 x 3 = 42 different mark forms. Then, the image processing unit 234 selects from the 42 mark forms the form corresponding to the baking method, baking temperature, and baking time corresponding to the characteristic experimental value, and forms the mark for that characteristic experimental value in the selected form.

[0244] The mark aspect may be, for example, the shape, color, size, etc. of the mark, or the color or width of the border of the mark, or may be something other than these. Note that the greater the number of different mark aspects, the lower the visibility or distinguishability of those marks. Therefore, it is preferable that the product of the number of non-utilized variables is a finite number that allows those marks to be distinguished. For example, the number of different mark aspects is 2 or more and 15 or less.

[0245] Thus, in the property display device 230 of this embodiment, when at least one non-utilized variable is associated with each of the experimental property values ​​of one or more tested compounds, the image processing unit 234 calculates the product of the number of cases of the at least one non-utilized variable as the number of types of mark forms. Then, for each of the experimental property values ​​of the one or more tested compounds, the image processing unit 234 selects a form from the number of types corresponding to the combination of data indicated by each of the at least one non-utilized variable associated with the experimental property value, and superimposes the mark of the selected form on the image element map Ma. For example, the at least one non-utilized variable indicates, as data, the baking temperature, baking time, and baking method used in baking the compound having the experimental property value.

[0246] This allows the user to visually and easily grasp the data of at least one non-utilized variable associated with each experimental characteristic value superimposed on the image element map Ma from the shape of the mark for that experimental characteristic value, thereby improving the efficiency of material search and appropriately supporting material development.

[0247] Furthermore, in the property display device 230 of this embodiment, each of the at least one unutilized variable indicates a process condition used to generate a compound having an experimental property value, or an attribute of the compound having an experimental property value.

[0248] This allows the user to visually and easily grasp the process conditions or attributes of a compound having an experimental characteristic value displayed superimposed on the image element map Ma from the shape of the mark for that experimental characteristic value.

[0249] In the characteristic display device 230 according to the present embodiment, the image processing unit 234 calculates a number between 2 and 15 as the number of types of mark forms.

[0250] For example, if the number of mark shape types were 16 or more, it would be difficult to distinguish between the shapes. Therefore, by limiting the number of types to 2 or more and 15 or less, the distinguishability of the shapes can be improved. As a result, the user can more easily grasp the data for at least one non-utilized variable associated with each characteristic experimental value superimposed and displayed on the image element map Ma. This further improves the efficiency of material search. In the above example, if the number of firing method cases is 2, the number of firing time cases is 7, and the number of firing temperature cases is 3, the image processing unit 234 determines 2 × 7 × 3 = 42 different mark shapes. In this case, since the number of mark shape types is 42, which is 16 or more, the image processing unit 234 may omit the firing temperature from the three non-utilized variables. As a result, the image processing unit 234 may determine 2 × 7 = 14 different mark shapes, thereby limiting the number of mark shape types to 15 or less.

[0251] FIG. 22A is a diagram showing another example of the map.

[0252] 22A , the image processing unit 234 may superimpose marks of characteristic experimental values ​​on the image element map Ma in a manner according to the analytical information corresponding to the characteristic experimental values. As described above, the analytical information indicates, for example, the crystalline structure of a compound. The types of crystalline structure include, for example, crystalline phase A and crystalline phase B. Furthermore, the analytical information is associated with the characteristic experimental values ​​of each compound in the experimental data of the experimental database 150.

[0253] For example, the display method acquisition unit 133 acquires second display method information from the input unit 110, which prompts a display according to the analytical information, and outputs the second display method information to the image processing unit 234. The experimental value acquisition unit 232 outputs the analytical information of each compound, along with the structure and characteristic experimental value of the compound, to the image processing unit 234. In accordance with the second display method information, the image processing unit 234 forms marks of the characteristic experimental values ​​in a manner according to the analytical information corresponding to the characteristic experimental values, and superimposes the marks on the image element map Ma, as shown in FIG. 22A . For example, the image processing unit 234 forms circular marks of the characteristic experimental values ​​of compounds having a crystal structure of crystalline phase A, and triangular marks of the characteristic experimental values ​​of compounds having a crystal structure of crystalline phase B.

[0254] FIG. 22B is a diagram showing another example of the map.

[0255] The analytical information may indicate the type of raw material used to synthesize the compound. ２ O (purity 99%), Li ２ In this case, the image processing unit 234 may superimpose the mark of the characteristic experimental value on the image element map Ma in a manner according to the type of raw material corresponding to the characteristic experimental value, as shown in FIG. 22B. For example, the image processing unit 234 may superimpose the mark of the characteristic experimental value on the image element map Ma in a manner according to the type of raw material corresponding to the characteristic experimental value. ２ The marks of the characteristic experimental values ​​of the compound synthesized from "LiO (purity 99%)" are formed in a circle, and the raw material "Li ２ The marks for the characteristic experimental values ​​of the compound synthesized from "C12H2O (purity 99.99%)" are formed into triangles.

[0256] FIG. 22C is a diagram showing another example of the map.

[0257] The above-mentioned analysis information may indicate whether or not raw materials used to synthesize the compound remain. In this case, as shown in FIG. 22C , the image processing unit 234 may superimpose marks of the characteristic experimental values ​​on the image element map Ma in a manner corresponding to whether or not the raw materials corresponding to the characteristic experimental values ​​remain. For example, the image processing unit 234 forms circular marks for the characteristic experimental values ​​of compounds with raw materials remaining (i.e., raw materials remaining: present), and forms triangular marks for the characteristic experimental values ​​of compounds with no raw materials remaining (i.e., raw materials remaining: absent).

[0258] Thus, in the characteristic display device 230 of this embodiment, the attributes of a compound are (a) the type of crystalline phase of the compound, (b) the type of raw material of the compound, or (c) whether or not the raw material of the compound remained when the compound was produced.

[0259] This allows the user to easily visually grasp the type of crystalline phase, the type of raw material, or whether or not any raw material remains of a compound having a characteristic experimental value superimposed on the image element map Ma from the shape of the mark for that characteristic experimental value.

[0260] [Processing Operation] FIG. 23 is a flowchart showing the processing operation of the display system 200 in this embodiment.

[0261] The display system 200 of this embodiment executes the processes of steps S111 to S113, similar to the display system 100 of embodiment 1A. Next, the display system 200 executes the processes of steps S121 to S124.

[0262] (Step S121) The experimental value acquisition unit 232 acquires the structure and experimental characteristic values ​​of each compound from the experimental database 150, and outputs them to the image processing unit 234. Note that if the structure and experimental characteristic values ​​of each compound are associated with process conditions, analysis information, and the like in the experimental database 150, the experimental value acquisition unit 232 may acquire the process conditions, analysis information, and the like together with the experimental characteristic values, and output them to the image processing unit 234.

[0263] (Step S122) As in step S113, the display method acquisition unit 133 acquires second display method information output from the input unit 110 in response to an input operation by the user to the input unit 110. This second display method information is information indicating a display method for the characteristic experiment values, for example, as shown in Figures 15 to 22C. Then, the display method acquisition unit 133 outputs the second display method information to the image processing unit 234.

[0264] (Step S123) The image processing unit 234 acquires the predicted property values ​​and structures of each of the multiple compounds from the predicted value acquisition unit 132 and receives first display method information from the display method acquisition unit 133. Then, the image processing unit 234 generates a map showing the predicted property values ​​of each of the multiple compounds according to the first display method information. The map may be an image element map Ma or an image map Mb. Furthermore, the image processing unit 234 acquires the experimental property values ​​and structures of each of the one or more compounds from the experimental value acquisition unit 232 and receives second display method information from the display method acquisition unit 133. Then, the image processing unit 234 superimposes the experimental property values ​​of each of the one or more compounds on the map according to the second display method information. This generates a map on which the experimental property values ​​of the one or more compounds are superimposed. The image processing unit 234 outputs an image including the map on which the experimental property values ​​of the one or more compounds are superimposed to the display unit 140.

[0265] (Step S124) The display unit 140 acquires an image from the image processing unit 234 and displays the image, that is, an image including a map on which the experimental characteristic values ​​of one or more compounds are superimposed.

[0266] By executing the processes of steps S111 to S113 and S121 to S124, the predicted and experimental property values ​​of the compound are displayed. This allows the experimental property values ​​to be overlaid on a map showing the predicted property values, allowing the user to properly grasp the overall picture of the experimental property values. Furthermore, the user can easily grasp the relationship between the predicted property values ​​and the experimental property values.

[0267] If no characteristic experimental value is acquired in step S121, the characteristic display device 230 executes the processes of steps S114 and S115 instead of steps S121 to S124, similar to the characteristic display device 130 of embodiment 1A.

[0268] (Variation of Embodiment 1B) The image processing unit 234 may generate an image map Mb as in Variation 1 of Embodiment 1A and superimpose the characteristic experimental value on the image map Mb. Note that the image map Mb is a map consisting of an arrangement of a plurality of image element maps Ma.

[0269] 24A and 24B are diagrams showing an example of the image map Mb in this modified example.

[0270] The image processing unit 234 superimposes, on any one of the image element maps Ma included in the image map Mb, the experimental characteristic values ​​of a compound having a structure corresponding to that image element map Ma. Here, the structures and experimental characteristic values ​​of each compound acquired by the experimental value acquiring unit 232 may include structures and experimental characteristic values ​​that do not correspond to any of the image element maps Ma in the image map Mb.

[0271] For example, as shown in FIG. 24A, the image map Mb corresponds to variables (M3, M3', M4, M4') = (La, Ga, Ti, Zr). Each image element map Ma included in this image map Mb corresponds to a combination of the respective option data of the discrete variables a and b. Here, for example, the composition formula of a compound "Li ０．７７ (La ０．５ Ga ０．５ ) ０．１５ (Ti ０．５ Zr ０．５ ) １．１９５ O ３ " and the characteristic experimental value are acquired by the experimental value acquisition unit 232. In this case, the discrete variable b in the composition formula of the compound is 0.195, and therefore the composition formula does not correspond to any of the image element maps Ma in the image maps Mb.

[0272] Therefore, in this modification, the image processing unit 234 represents the composition formula of the compound acquired by the experimental value acquisition unit 232 and the composition formula corresponding to each position on the image map Mb as vectors. The composition formula vector acquired by the experimental value acquisition unit 232 is hereinafter referred to as the first vector, and the composition formula vector corresponding to each position on the image map Mb is hereinafter referred to as the second vector. These vectors are defined by, for example, six variables (M3, M3', M4, M4', a, b, x, y). The image processing unit 234 calculates the norm of the difference between the first vector and the second vector at each position on the image map Mb as a distance. This distance may be a normalized distance as described above. Then, as shown in FIG. 24A , the image processing unit 234 superimposes the characteristic experimental value of the compound acquired by the experimental value acquisition unit 232 on the position 1700 on the image map Mb corresponding to the shortest distance among the calculated distances. The position 1700 is a position on one image element map Ma included in the image map Mb.

[0273] Furthermore, the image processing unit 234 may superimpose the characteristic experimental value at a position on the image map Mb that corresponds to the shortest distance among the calculated distances that is equal to or less than a threshold value. The threshold value is, for example, 0.01. Alternatively, the image processing unit 234 may superimpose the characteristic experimental value at a position on the image map Mb that corresponds to a distance equal to or less than a threshold value among the calculated distances. If there are multiple positions on the image map Mb that correspond to distances equal to or less than the threshold value, the image processing unit 234 may superimpose the characteristic experimental value at these multiple positions.

[0274] Furthermore, when there are a plurality of shortest distances among the calculated distances, the image processing unit 234 may superimpose the characteristic experimental value at a position 1701 on the image map Mb corresponding to each of the plurality of shortest distances, as shown in FIG. 24B. For example, when the composition formula of a compound is "Li １．４７５ (La ０．５ Ga ０．５ ) ０．１７５ (Ti ０．５ Zr ０．５ ) １．０ O ３" and the characteristic experimental value are acquired by the experimental value acquisition unit 232. In this case, the discrete variable a in the composition formula of the compound is 0.175, and therefore the composition formula does not correspond to any of the image element maps Ma in the image maps Mb.

[0275] In this case, the image processing unit 234 calculates, as a distance, the norm of the difference between the first vector corresponding to the composition formula and the second vector at each position on the image map Mb, as described above. Then, in the example of FIG. 24B , the image processing unit 234 identifies two positions 1701 on the image map Mb corresponding to the shortest distance among the calculated distances. Note that these two positions 1701 are included in two adjacent image element maps Ma. The image processing unit 234 superimposes the characteristic experimental value of the compound acquired by the experimental value acquisition unit 232 onto the two positions 1701. Note that the distance corresponding to the position 1700 is equal to or less than the above-mentioned threshold value "0.01," and the distance corresponding to the two positions 1701 is longer than the threshold value. Therefore, when the condition of being equal to or less than the threshold is applied, i.e., when the characteristic experimental value is superimposed on a position on the image map Mb corresponding to a distance equal to or less than the threshold, even if the characteristic experimental value is superimposed on the position 1700, the characteristic experimental value is not superimposed on the two positions 1701.

[0276] Thus, in the characteristic display device 230 of this embodiment, the image map Mb includes a plurality of image element maps Ma arranged in a matrix along the first and second coordinate axes, and each of the plurality of image element maps Ma has a third and fourth coordinate axis. The image processing unit 234 then associates the first, second, third, and fourth coordinate axes with the first, second, third, and fourth variables, respectively, of the plurality of variables. For each of the plurality of compounds, the image processing unit 234 identifies an image element map Ma among the plurality of image element maps Ma associated with the values ​​of the first and second variables used to represent the compound's structure. Next, the image processing unit 234 maps the predicted characteristic value of the compound to a position on the identified image element map Ma corresponding to the values ​​of the third and fourth variables used to represent the compound's structure. For example, the first, second, third, and fourth variables are the variables a, b, x, and y described above.

[0277] This allows maps such as image map Mb to show predicted property values ​​for multiple compound structures, each represented by four variables, and to clearly display predicted property values ​​for compounds over a wide range, thereby improving the efficiency of materials search and appropriately supporting materials development.

[0278] Furthermore, in the characteristic display device 230 of this embodiment, when there is no image element map Ma among the multiple image element maps Ma for each of one or more experimental compounds that is associated with the values ​​of the first and second variables used to represent the structure of that compound, the image processing unit 234 identifies, instead of that image element map Ma, image element maps Ma associated with values ​​that are closest to the values ​​of the first and second variables. Then, the image processing unit 234 superimposes the characteristic experimental value of that compound on the identified image element map Ma at a position that corresponds to the value of the third and fourth variables used to represent the structure of that compound.

[0279] For example, since the first variable and the second variable are discrete variables, there may be cases where there are no image element maps Ma associated with the values ​​of the first variable and the second variable. However, in this embodiment, the image element maps Ma associated with the values ​​closest to those variables are identified, and the characteristic experimental value is superimposed on that image element map Ma. Therefore, it is possible to prevent the characteristic experimental value from being hidden from the image element map Ma due to the discrete variables.

[0280] As described above, in the first embodiment including the embodiments 1A and 1B and their modifications, the display method of the predicted characteristic values ​​can be appropriately set by the first display method information. Also, in the embodiment 1B and its modifications, the display method of the experimental characteristic values ​​can be appropriately set by the second display method information.

[0281] In this embodiment, the display method acquisition unit 133 acquires the first display method information generated in response to an input operation by the user on the input unit 110 .

[0282] This allows the user to arbitrarily set the display method for the predicted characteristic values, thereby improving convenience. The same applies to the second display method information, where the user can arbitrarily set the display method for the experimental characteristic values, thereby improving convenience.

[0283] On the other hand, the display method acquisition unit 133 may acquire the first display method information by determining a display method for the predicted property values ​​based on the predicted property values ​​of each of the multiple compounds acquired by the predicted value acquisition unit 132. For example, the display method acquisition unit 133 determines, as the display method for the predicted property values, to generate an image element map Ma by associating predetermined colors or color shadings for indicating the maximum and minimum values ​​of the predicted property values ​​on the image element map Ma with the maximum and minimum values ​​of the predicted property values ​​of the multiple compounds acquired by the predicted value acquisition unit 132.

[0284] As a result, the display method for the predicted property value is determined based on the predicted property value, so that a display suited to the predicted property value can be performed, thereby improving the efficiency of material search.

[0285] In a specific example, the maximum value of the predicted property values ​​of the multiple compounds acquired by the predicted value acquisition unit 132 is 3.5, and the minimum value is 0.0. In this case, the display method acquisition unit 133 assigns a predetermined maximum shading to the maximum value of the predicted property value, "3.5," and assigns a predetermined minimum shading to the minimum value of the predicted property value, "0.0." In other words, the display method acquisition unit 133 determines this relationship between the shading and the predicted property value as the display method for the predicted property value. As a result, the display method acquisition unit 133 acquires first display method information indicating the display method. Next, for example, by updating the predictor in the predictor database 120, the predicted value acquisition unit 132 acquires new predicted property values ​​for the multiple compounds. In this case, for example, the maximum value of the predicted property values ​​of the multiple compounds newly acquired by the predicted value acquisition unit 132 is 1.5, and the minimum value is 1.0. In this case, the display method acquisition unit 133 assigns the predetermined maximum shading to the maximum value of the predicted property value, "1.5," and the predetermined minimum shading to the minimum value of the predicted property value, "1.0." In other words, the display method acquisition unit 133 determines this new relationship between shading and predicted property value as the display method for the predicted property value. As a result, the display method acquisition unit 133 acquires first display method information indicating the new display method. The image processing unit 234 generates an image element map Ma indicating the predicted property value in accordance with the first display method information indicating the display method determined based on the predicted property value. As a result, regardless of whether the range of the acquired predicted property value is wide or narrow, or whether the predicted property value within that range is large or small, the predicted property value is displayed with a shading according to that range. Therefore, the user can easily grasp the predicted property value from the shading, thereby improving the efficiency of material search.

[0286] (Embodiment 2A) Similar to Embodiment 1A, the display system of this embodiment displays predicted property values ​​for each of a plurality of compounds in the form of a map, and superimposes one or more candidate points on the map. The candidate points indicate the configuration of candidate compounds (i.e., candidate compounds) to be used in future experiments. Note that, among the components of this embodiment, the same components as those of Embodiments 1A and 1B are designated by the same reference numerals as those of Embodiments 1A and 1B, and detailed description thereof will be omitted.

[0287] [Configuration of Display System 300] Fig. 25 is a block diagram showing an example of the configuration of a display system 300 according to this embodiment. The display system 300 shown in Fig. 25 includes an input unit 110, a predictor database 120, a characteristic display device 330, a display unit 140, and an experiment database 150. The characteristic display device 330 is an example of an information display device.

[0288] The characteristic display device 330 in this embodiment determines candidate points based on the acquired predicted characteristic values ​​of the compound or based on the predicted characteristic values ​​and the positions of the experimental characteristic values, and superimposes the candidate points on the image element map Ma. The positions of the experimental characteristic values ​​are positions on the image element map Ma corresponding to the structure of the experimentally obtained compound from which the experimental characteristic values ​​were obtained. The characteristic display device 330 then displays an image including the image element map Ma with the superimposed candidate points on the display unit 140. The characteristic display device 330 includes a search range acquisition unit 131, a predicted value acquisition unit 132, a display method acquisition unit 133, an experimental value acquisition unit 232, a candidate point determination unit 331, and an image processing unit 334. The characteristic display device 330 may be composed of a processor such as a CPU and a memory. In this case, the processor functions as the characteristic display device 330 by executing a computer program stored in the memory, for example. The memory may be volatile or nonvolatile, or may be composed of a volatile memory and a nonvolatile memory.

[0289] [Predicted Value Acquisition Unit 132] In this embodiment, the predicted value acquisition unit 132 associates the structure of each of a plurality of compounds with the predicted property value acquired for that structure and outputs the associated structure to the image processing unit 334, and also outputs the structure and predicted property value to the candidate point determination unit 331. Note that the structure of a compound may be the composition formula of the compound, or, if the search range includes process conditions for producing the compound, the structure of the compound may be the composition formula of the compound and the process conditions. The process conditions include, for example, a firing method, a firing time, a firing temperature, etc.

[0290] [Experimental Value Acquisition Unit 232] The experimental value acquisition unit 232 in this embodiment acquires the structure and characteristic experimental values ​​of each of one or more compounds that have been tested from the experiment database 150. Then, the experimental value acquisition unit 232 outputs the acquired structure and characteristic experimental values ​​for each of the one or more compounds that have been tested to the candidate point determination unit 331.

[0291] [Display method acquisition unit 133] The display method acquisition unit 133 acquires third display method information indicating a display method of the candidate points from the input unit 110. Then, the display method acquisition unit 133 outputs the third display method information to the candidate point determination unit 331 and the image processing unit 334.

[0292] [Candidate Point Determining Unit 331] The candidate point determining unit 331 acquires the compound structures and predicted property values ​​from the predicted value acquiring unit 132, and acquires the compound structures and experimental property values ​​from the experimental value acquiring unit 232. Furthermore, the candidate point determining unit 331 acquires third display method information from the display method acquiring unit 133. Then, the candidate point determining unit 331 determines one or more positions from all positions on the image element map Ma as candidate points according to the third display method information. At this time, the candidate point determining unit 331 determines candidate points based on the predicted property values ​​indicated at each position on the image element map Ma. Alternatively, the candidate point determining unit 331 determines candidate points based on the predicted property values ​​indicated at each position and the positions on the image element map Ma corresponding to the respective structures of one or more experimentally tested compounds, i.e., the positions of the experimental property values.

[0293] For example, the candidate point determination unit 331 determines, as a candidate point, a first position on the image element map Ma where a predicted characteristic value equal to or greater than a threshold is indicated. Alternatively, the candidate point determination unit 331 determines, as a candidate point, a second position on the image element map Ma where a predicted characteristic value equal to or less than a threshold is indicated. Alternatively, the candidate point determination unit 331 determines, as a candidate point, a third position on the image element map Ma where a predicted characteristic value whose difference from a target value is equal to or less than a threshold is indicated. The target value may be a value set in response to an input operation by a user to the input unit 110. In this case, the candidate point determination unit 331 acquires the target value from the input unit 110 via the display method acquisition unit 133 and determines a candidate point using the target value.

[0294] In the above example, the candidate point determiner 331 determines a first position at which a predicted characteristic value equal to or greater than the threshold is indicated as a candidate point. However, it may also be possible to determine, as a candidate point, first positions at which the top K predicted characteristic values ​​are indicated among all predicted characteristic values ​​on the image element map Ma. K is a predetermined integer equal to or greater than 1, and the top K predicted characteristic values ​​are greater than any other predicted characteristic values. Furthermore, the candidate point determiner 331 determines, as a candidate point, a second position at which a predicted characteristic value equal to or less than the threshold is indicated. However, it may also be possible to determine, as a candidate point, second positions at which the top L predicted characteristic values ​​are indicated among all predicted characteristic values ​​equal to or less than the threshold. L is a predetermined integer equal to or greater than 1, and the top L predicted characteristic values ​​are greater than any other predicted characteristic values ​​equal to or less than the threshold. Furthermore, the candidate point determiner 331 determines, as a candidate point, a third position at which a predicted characteristic value whose difference from the target value is equal to or less than the threshold is indicated. However, it may also be possible to determine, as a candidate point, third positions at which the bottom J predicted characteristic values ​​whose difference from the target value is indicated among all predicted characteristic values ​​on the image element map Ma. Here, J is a predetermined integer equal to or greater than 1, and the difference between the lowest J predicted characteristic values ​​and the target value is smaller than any other predicted characteristic value.

[0295] The candidate point determination unit 331 may also determine, as candidate points, two or more first positions that are separated from each other by a predetermined distance or more among the multiple first positions. Similarly, the candidate point determination unit 331 may also determine, as candidate points, two or more positions that are separated from each other by a predetermined distance or more among the multiple second positions or multiple third positions. Furthermore, when the structures and experimental characteristic values ​​of two experimentally tested compounds have been obtained, the candidate point determination unit 331 may determine, as candidate points, positions that are midpoints between the positions of the two experimental characteristic values ​​on the image element map Ma. Furthermore, the candidate point determination unit 331 may also determine, as candidate points, first positions that are separated from any of the experimental characteristic values ​​by a predetermined distance or more among the multiple first positions. Similarly, the candidate point determination unit 331 may determine, as candidate points, positions that are separated from any of the experimental characteristic values ​​by a predetermined distance or more among the multiple second positions or multiple third positions.

[0296] [Image Processing Unit 334] The image processing unit 334 generates an image element map Ma in the same manner as in embodiment 1A. Furthermore, the image processing unit 334 acquires third display method information from the display method acquisition unit 133, and superimposes one or more candidate points determined by the candidate point determination unit 331 on the image element map Ma in accordance with the third display method information. That is, the image processing unit 334 superimposes marks indicating each of the one or more candidate points on the image element map Ma. Then, the image processing unit 334 outputs an image including the image element map Ma on which the one or more candidate points are superimposed to the display unit 140.

[0297] [Display Unit 140] The display unit 140 acquires an image from the image processing unit 334, and displays the image, that is, the image including the image element map Ma on which the one or more candidate points are superimposed.

[0298] [Specific Example of Map] FIG. 26 is a diagram showing an example of an image element map Ma on which candidate points are superimposed.

[0299] The image processing unit 334 superimposes the candidate points as marks such as star-shaped marks on the image element map Ma of embodiment 1A in accordance with the third display method information. That is, the candidate point determination unit 331 determines one or more candidate points from all positions on the image element map Ma in accordance with the third display method information. Then, the image processing unit 334 superimposes star-shaped marks on the one or more candidate points.

[0300] As described above, the property display device 330 in this embodiment includes a predicted value acquisition unit 132 and an image processing unit 334. The predicted value acquisition unit 132 acquires predicted property values ​​for each of a plurality of compounds. The image processing unit 334 generates an image element map Ma indicating the predicted property values ​​of each of the plurality of compounds at positions corresponding to the respective structures of the compounds. Next, the image processing unit 334 superimposes candidate points indicating the respective structures of one or more candidate compounds on the image element map Ma at positions corresponding to the structures of the candidate compounds. The image processing unit 334 then generates and outputs an image including the image element map Ma on which one or more candidate points are superimposed. Here, the candidate compounds are compounds that are considered experimental candidates among a plurality of compounds each having a predicted property value indicated in the image element map Ma. For example, the image is output to and displayed on the display unit 140.

[0301] This allows the user to easily grasp the relationship between each predicted property value in the image element map Ma and one or more candidate points, thereby improving the efficiency of material search. In other words, the user can easily find promising candidate points for the next experiment from the image element map Ma in the image displayed on the display unit 140. As a result, the efficiency of material search can be improved, and material development can be appropriately supported.

[0302] The property display device 330 does not necessarily have to include the predicted value acquisition unit 132. The image processing unit 334 generates the image element map Ma, but may acquire the image element map Ma without generating it. That is, the image processing unit 334 acquires an image element map Ma indicating predicted property values ​​of multiple compounds at positions corresponding to the respective structures of the compounds. The image processing unit 334 then outputs an image including the image element map Ma, which is generated by superimposing candidate points indicating the respective structures of one or more candidate compounds on the image element map Ma at positions corresponding to the structures of the candidate compounds. As described above, the candidate compounds are compounds that are considered as experimental candidates among multiple compounds each having a predicted property value indicated in the image element map Ma. Even with this property display device 330, the same effects as those described above can be obtained.

[0303] The characteristic display device 330 in this embodiment also includes a display method acquisition unit 133 that acquires third display method information that indicates a display method for the candidate points. The image processing unit 334 superimposes one or more candidate points on the image element map Ma in accordance with the third display method information.

[0304] As a result, the candidate points are superimposed on the image element map Ma in accordance with the display method indicated by the third display method information, and therefore, depending on the setting of the display method, the candidate points can be displayed in a manner that suits the purpose of the user's material search, thereby improving the efficiency of material search.

[0305] The property display device 330 in this embodiment also includes a candidate point determination unit 331 that determines one or more candidate points based on the predicted property values ​​of each of the multiple compounds. The image processing unit 334 superimposes the one or more candidate points determined by the candidate point determination unit 331 on the image element map Ma.

[0306] This allows candidate points to be determined based on the predicted property values ​​of each of the multiple compounds, so that the positions showing the best predicted property values, such as the maximum predicted property value, the minimum predicted property value, or the predicted property value closest to the target value, can be determined as candidate points, thereby further improving the efficiency of material search.

[0307] The property display device 330 of this embodiment also includes a search range acquisition unit 131 that acquires a search range. This search range acquisition unit 131 acquires multiple variables used to express the structure of a compound and, for each of the multiple variables, multiple option data indicating possible values ​​or elements for the variable. The predicted value acquisition unit 132 acquires, for each combination of option data obtained by selecting one option data from the multiple option data for each of the multiple variables, a predicted property value of a compound having a structure corresponding to that combination. In other words, the predicted value acquisition unit 132 acquires, for each of the above combinations, a predicted property value of a compound having a structure corresponding to that combination using a predetermined algorithm.

[0308] This allows appropriate predicted property values ​​to be obtained. That is, the property values ​​of multiple compounds with different compositions can be appropriately predicted. Furthermore, when the predetermined algorithm is updated based on experiments on compounds that are conducted from time to time, the prediction accuracy of the property values ​​can be improved.

[0309] FIG. 27 is a diagram showing another example of the image element map Ma on which candidate points are superimposed.

[0310] The candidate point determination unit 331 may determine one or more candidate points from a plurality of points (hereinafter also referred to as grid points) in the image element map Ma represented by combinations of variables x and y according to the third display method information. The grid points are, for example, 11 x 11 points represented by variables (x, y) = (0.0, 0.0), (0.0, 0.1), (0.0, 0.2), ..., (1.0, 0.8), (1.0, 0.9), (1.0, 1.0). The image processing unit 334 then superimposes a mark on the determined one or more candidate points. In other words, the candidate point determination unit 331 extracts candidate points corresponding to the structure (i.e., composition formula) of a predetermined compound. Alternatively, it can be said that the candidate point determination unit 331 limits the candidate points to grid points.

[0311] For example, as shown in Figure 27, of the 11 x 11 grid points, two grid points with variables (x, y) = (0.2, 0.2) and (0.9, 0.1) are determined as candidate points, and marks indicating the candidate points are superimposed on the image element map Ma.

[0312] Thus, in the characteristic display device 330 of this embodiment, the third display method information indicates, as a display method for candidate points, that candidate points are superimposed only at a plurality of predetermined positions on the image element map Ma.

[0313] This makes it possible to prevent points that are difficult to use in experiments on the image element map Ma, such as points with variables (x, y) = (0.001, 0.002), or points that the user does not intend, from being determined and displayed as candidate points.

[0314] FIG. 28 is a diagram showing another example of the image element map Ma on which candidate points are superimposed.

[0315] The candidate point determination unit 331 may limit the number of candidate points to be determined to n or less (n is an integer equal to or greater than 1) according to the third display method information. In the example of FIG. 28 , n=5. For example, the candidate point determination unit 331 determines (n+1) or more tentative candidate points on the image element map Ma and identifies the priorities of these tentative candidate points. Then, the candidate point determination unit 331 determines the top n tentative candidate points with the highest priorities from the plurality of tentative candidate points as candidate points. The tentative candidate points are, for example, points located at the first position, second position, or third position described above. For example, if each of the (n+1) or more tentative candidate points is a first position, the priority may be the predicted characteristic value at the first position. For example, if each of the (n+1) or more tentative candidate points is a second position, the priority may be the predicted characteristic value at the second position. Furthermore, for example, if each of the (n+1) or more tentative candidate points is a third position, the priority may be the reciprocal of the difference between the predicted characteristic value and the target value at the third position, or the shortest distance from the position of each experimental characteristic value to the tentative candidate point.

[0316] The image processing unit 334 superimposes the n or less candidate points determined in this manner on the image element map Ma and displays them on the display unit 140. This allows the user to easily grasp the candidate points that are promising experimental candidates.

[0317] That is, in the characteristic display device 330 of this embodiment, the third display method information indicates the upper limit number of candidate points to be superimposed on the image element map Ma as a display method of the candidate points. In this case, the image processing unit 334 superimposes one or more candidate points on the image element map Ma, the number of which is equal to or less than the upper limit.

[0318] This limits the number of candidate points superimposed on the image element map Ma even when there are many points on the image element map Ma that can be candidate points. Therefore, for example, it is possible to narrow down the many points to only promising points as candidate points and present them to the user, thereby improving the efficiency of material search.

[0319] FIG. 29 is a diagram showing another example of the image element map Ma on which candidate points are superimposed.

[0320] The candidate point determination unit 331 may determine candidate points superimposed on the image element map Ma in accordance with the third display method information so that the linear distance between the candidate points is equal to or greater than a predetermined distance, as shown in FIG. 29 . The linear distance between candidate points is the linear distance between the positions on the image element map Ma indicated by the continuous variables (x, y) of the two candidate points. The predetermined distance is, for example, three times the step width (i.e., one step) of the continuous variables x and y. In other words, the predetermined distance can also be considered three steps. The predetermined distance is also referred to as the first minimum separation distance.

[0321] For example, the candidate point determination unit 331 determines, from among multiple tentative candidate points on the image element map Ma, the tentative candidate point that exhibits the best predicted characteristic value as the first candidate point. The tentative candidate point that exhibits the best predicted characteristic value may be the tentative candidate point with the highest priority. For example, if multiple tentative candidate points are each at the first position, the tentative candidate point that exhibits the best predicted characteristic value is the tentative candidate point that exhibits the largest predicted characteristic value among the multiple tentative candidate points.

[0322] Next, the candidate point determination unit 331 deletes tentative candidate points that are less than a predetermined distance from the first candidate point, and determines, from the one or more remaining tentative candidate points, the tentative candidate point that shows the next best characteristic prediction value after the first candidate point as the second candidate point. The candidate point determination unit 331 repeats this process until all tentative candidate points are deleted, thereby determining two or more candidate points. The straight-line distance between these candidate points is equal to or greater than the predetermined distance.

[0323] Alternatively, the candidate point determination unit 331 randomly determines two or more tentative candidate points from a plurality of tentative candidate points as candidate points, and moves one of the candidate points to a tentative candidate point at a different position from the candidate point so that the straight-line distance between the two candidate points is equal to or greater than a predetermined distance. In other words, the candidate point determination unit 331 moves one of the candidate points. By sequentially moving the candidate points in this manner, the candidate point determination unit 331 sets the straight-line distance between each of the two or more candidate points to be equal to or greater than the predetermined distance.

[0324] The image processing unit 334 superimposes the plurality of candidate points that are determined by the candidate point determination unit 331 as described above and that are spaced apart by a predetermined distance or more on the image element map Ma.

[0325] In this manner, in the characteristic display device 330 of this embodiment, the third display method information indicates the predetermined first minimum separation distance between the plurality of candidate points as the display method of the candidate points. In this case, the image processing unit 334 superimposes the plurality of candidate points that are separated from each other by at least the first minimum separation distance on the image element map Ma.

[0326] This allows the user to easily identify candidate points that have a certain difference in the compound formula (i.e., composition). It also prevents candidate points from being crowded together, improving visibility.

[0327] FIG. 30 is a diagram showing another example of the image element map Ma on which candidate points are superimposed.

[0328] The candidate point determiner 331 determines a plurality of candidate points and an experimental design according to the third display method information, as shown in FIG. 30 , for example. The experimental design is information indicating which of the plurality of candidate points should be tested and in what order. The candidate point determiner 331 outputs information indicating the determined plurality of candidate points and the experimental design to the image processor 334. For example, the candidate point determiner 331 determines a plurality of candidate points as described above, and then determines n candidate points from the plurality of candidate points in descending order of priority as candidate points to be included in the experimental design (hereinafter referred to as design candidate points). For example, n = 5. The candidate point determiner 331 then assigns experiment numbers to each of the n design candidate points in descending order of priority. For example, the candidate point determiner 331 assigns experiment number "1" to the design candidate point with the highest priority among the n design candidate points, and experiment number "2" to the design candidate point with the second highest priority. This generates an experimental design that indicates n design candidate points and the experiment numbers of these design candidate points.

[0329] The image processing unit 334 acquires information indicating a plurality of candidate points and an experimental design from the candidate point determination unit 331. The image processing unit 334 then superimposes the plurality of candidate points on the image element map Ma. Furthermore, the image processing unit 334 assigns the experiment number of each of the n candidate design points included in the experimental design to the image element map Ma, and further superimposes an arrow pointing from one candidate design point to the candidate design point of the next experiment number on the image element map Ma. Such superimposition of the arrows and the like is performed in accordance with the third display method information.

[0330] In a specific example, the image processing unit 334 assigns experiment numbers (1), (2), (3), (4), and (5) to five of the multiple candidate points. The image processing unit 334 then superimposes an arrow pointing from the candidate point of experiment number (1) to the candidate point of experiment number (2) and an arrow pointing from the candidate point of experiment number (2) to the candidate point of experiment number (3). Furthermore, the image processing unit 334 superimposes an arrow pointing from the candidate point of experiment number (3) to the candidate point of experiment number (4) and an arrow pointing from the candidate point of experiment number (4) to the candidate point of experiment number (5). The image processing unit 334 displays an image including the image element map Ma to which the experiment numbers and arrows have been assigned on the display unit 140. Therefore, an experimental plan indicating that experiments should be conducted in the following order is presented to the user from the display unit 140: candidate design point for experiment number (1), candidate design point for experiment number (2), candidate design point for experiment number (3), candidate design point for experiment number (4), and candidate design point for experiment number (5).

[0331] Thus, in the characteristic display device 330 of this embodiment, the third display method information indicates that the order of experiments on candidate compounds corresponding to each of one or more candidate points is to be displayed as an experimental plan as a display method for the candidate points.

[0332] This allows the user to know in what order experiments should be conducted on the candidate points when there are multiple candidate points. Therefore, by conducting experiments on the candidate points in that order, the user can efficiently conduct the experiments. In other words, the efficiency of material search can be improved, and material development can be appropriately supported. In the example of FIG. 30 , the experimental plan is indicated by experiment numbers and arrows, but the present invention is not limited to this and may be indicated in other ways.

[0333] FIG. 31 is a diagram showing another example of the image element map Ma on which candidate points are superimposed.

[0334] The candidate point determiner 331 may classify multiple candidate points into candidate points whose predicted characteristic values ​​are equal to or greater than a reference value and candidate points whose predicted characteristic values ​​are less than the reference value, based on the third display method information, as shown in FIG. 31 . The reference value may be, for example, the average of the predicted characteristic values ​​at each position on the image element map Ma, the median of the predicted characteristic values, or any other arbitrary value, but is not limited to these. The candidate point determiner 331 acquires third display method information related to the reference value from the input unit 110 via the display method acquirer 133 and determines the reference value based on the third display method information. Then, for each of the multiple candidate points, the candidate point determiner 331 determines whether the predicted characteristic value indicated at that candidate point is equal to or greater than the reference value. If the candidate point determines that the predicted characteristic value is equal to or greater than the reference value, the candidate point determiner 331 assigns a flag indicating, for example, 1 to that candidate point. On the other hand, if the candidate point determiner 331 determines that the predicted characteristic value indicated at that candidate point is less than the reference value, the candidate point determiner 331 assigns a flag indicating, for example, 0 to that candidate point. Then, the candidate point determination unit 331 outputs the plurality of candidate points and flags to the image processing unit 334 .

[0335] When the image processing unit 334 acquires the plurality of candidate points and flags from the candidate point determination unit 331, it assigns, for example, a star-shaped mark to candidate points to which a flag indicating 1 has been assigned, in accordance with the third display method information, as shown in Fig. 31 . Furthermore, the image processing unit 334 assigns, for example, an inverted triangle mark to candidate points to which a flag indicating 0 has been assigned. Then, the image processing unit 334 superimposes the mark of the form assigned to the candidate point on the image element map Ma, and displays an image including the image element map Ma with the superimposed mark on the display unit 140. As a result, the candidate points indicated by the marks of each form are displayed superimposed on the image element map Ma.

[0336] Therefore, candidate points whose predicted characteristic values ​​are equal to or greater than the reference value can be highlighted more than candidate points whose predicted characteristic values ​​are less than the reference value. In addition, multiple candidate points that require attention can be simultaneously identified while grasping the entire search range.

[0337] Note that if the reference value is an arbitrary value, for example, equal to the tenth largest predicted characteristic value in the image element map Ma, the candidate points showing the largest predicted characteristic value through the candidate point showing the tenth largest predicted characteristic value can be highlighted. Alternatively, conversely to the above example, candidate points showing predicted characteristic values ​​less than the reference value may be highlighted more than candidate points showing predicted characteristic values ​​equal to or greater than the reference value.

[0338] In this manner, in the property display device 330 of this embodiment, the third display method information indicates, as a display method of candidate points, that among one or more candidate points, a candidate point whose predicted property value indicated at the position on the image element map Ma on which the candidate point is superimposed is equal to or greater than a reference value, and is displayed in a more emphasized manner than the remaining candidate points. For example, in the property display device 330 of this embodiment, the reference value is the average or median of the predicted property values ​​of each of the multiple compounds, or a value specified by the user.

[0339] This allows the user to easily find candidate points that show predicted property values ​​equal to or greater than the reference value from among one or more candidate points superimposed on the image element map Ma, thereby improving the efficiency of material search.

[0340] FIG. 32 is a diagram showing another example of the image element map Ma on which candidate points are superimposed.

[0341] The candidate point determination unit 331 may classify multiple candidate points into three groups according to the third display method information, as shown in FIG. 32 . In the example of FIG. 32 , the three groups are candidate point group 1, candidate point group 2, and candidate point group 3. Classification into these groups may be performed by comparing predicted characteristic values ​​with reference values, as in the example of FIG. 31 . For example, if the predicted characteristic value of a candidate point is equal to or greater than a first reference value, the candidate point determination unit 331 classifies the candidate point into candidate point group 1. Furthermore, if the predicted characteristic value of a candidate point is less than the first reference value and equal to or greater than a second reference value, the candidate point determination unit 331 classifies the candidate point into candidate point group 2. Furthermore, if the predicted characteristic value of a candidate point is less than the second reference value, the candidate point determination unit 331 classifies the candidate point into candidate point group 3. The first and second reference values ​​may be, for example, the average value of the predicted characteristic values ​​at each position on the image element map Ma, the median value of the predicted characteristic values, or any other arbitrary value, but are not limited to these. The first reference value is greater than the second reference value. In a specific example, the first reference value is 2.5 eV, and the second reference value is 1.5 eV.

[0342] Alternatively, the above-described priority may be used instead of the predicted characteristic value of the candidate point to classify the candidate points into each group. In this case, the priority used may be the predicted characteristic value or the reciprocal of the difference between the predicted characteristic value and the target value. For example, if the priority of a candidate point is equal to or greater than the first priority reference value, the candidate point determination unit 331 classifies the candidate point into candidate point group 1. If the priority of a candidate point is less than the first priority reference value and equal to or greater than the second priority reference value, the candidate point determination unit 331 classifies the candidate point into candidate point group 2. If the priority of a candidate point is less than the second priority reference value, the candidate point determination unit 331 classifies the candidate point into candidate point group 3. The first priority reference value and the second priority reference value are reference values ​​compared with the priority, and the first priority reference value is greater than the second priority reference value. The first priority reference value and the second priority reference value may be collectively referred to as priority reference values ​​hereinafter. The priority reference value may be, but is not limited to, the average value of the priorities of the candidate points, the median value of the priorities, or any other arbitrary value.

[0343] After classifying the multiple candidate points into three groups in this way, the candidate point determination unit 331 associates the multiple candidate points with group identification information indicating the group to which the candidate points belong and outputs them to the image processing unit 334.

[0344] 32 , the image processing unit 334 assigns, for example, a star-shaped mark to the candidate points associated with the group identification information indicating candidate point group 1. The image processing unit 334 also assigns, for example, an inverted triangle mark to the candidate points associated with the group identification information indicating candidate point group 2. The image processing unit 334 also assigns, for example, a square mark to the candidate points associated with the group identification information indicating candidate point group 3. The assignment of these marks is performed in accordance with the third display method information.

[0345] Then, the image processing unit 334 superimposes marks of the forms assigned to the candidate points on the image element map Ma, and displays an image including the image element map Ma with the superimposed marks on the display unit 140. As a result, the candidate points indicated by the marks of each form are displayed superimposed on the image element map Ma.

[0346] In this manner, in the characteristic display device 330 of this embodiment, the third display method information indicates, as a display method of the candidate points, that each of one or more candidate points is classified into one of a plurality of groups according to the priority of the candidate point, and a mark of a type associated with that group is superimposed on the image element map Ma. The priority of the candidate point may be, for example, a characteristic prediction value indicated at the candidate point.

[0347] This allows the user to easily understand the group to which each candidate point superimposed on the image element map Ma belongs, i.e., the characteristics of the candidate points belonging to that group, such as whether their priority is high, medium, or low.

[0348] [Processing Operation] FIG. 33 is a flowchart showing the processing operation of the display system 300 in this embodiment.

[0349] Display system 300 in this embodiment executes the processes of steps S111 and S112, similar to display system 100 in embodiment 1A, and executes step S121, similar to display system 200 in embodiment 1B. Next, display system 300 executes the processes of steps S131 to S134.

[0350] (Step S131) ​​As in step S113, the display method acquisition unit 133 acquires third display method information output from the input unit 110 in response to an input operation by the user to the input unit 110. This third display method information is information indicating a method of displaying candidate points, for example, as shown in Figures 26 to 32. Then, the display method acquisition unit 133 outputs the third display method information to the candidate point determination unit 331 and the image processing unit 334.

[0351] (Step S132) The candidate point determination unit 331 acquires the structure and predicted property values ​​of each of the multiple compounds from the predicted value acquisition unit 132, and acquires the structure and experimental property values ​​of each of the one or more compounds from the experimental value acquisition unit 232. Furthermore, the candidate point determination unit 331 acquires third display method information from the display method acquisition unit 133. Then, in accordance with the third display method information, the candidate point determination unit 331 determines one or more candidate points based on the structure and predicted property values ​​of each of the multiple compounds and the structure and experimental property values ​​of each of the one or more compounds.

[0352] (Step S133) The image processing unit 334 acquires, for each of the multiple compounds, the structure and predicted property values ​​of that compound from the predicted value acquisition unit 132. Furthermore, the image processing unit 334 acquires one or more candidate points from the candidate point determination unit 331. The image processing unit 334 also acquires third display method information from the display method acquisition unit 133. Note that, as in embodiment 1A, the image processing unit 334 may also receive first display method information indicating a display method for the predicted property values ​​from the display method acquisition unit 133.

[0353] The image processing unit 334 generates an image element map Ma based on the acquired structures and predicted property values ​​of each of the multiple compounds. At this time, if the image processing unit 334 has received first display method information, it may generate the image element map Ma in accordance with the first display method information. Furthermore, the image processing unit 334 superimposes one or more candidate points acquired from the candidate point determination unit 331 on the image element map Ma in accordance with third display method information. That is, the image processing unit 334 superimposes marks indicating one or more candidate points on the image element map Ma. The image processing unit 334 outputs an image including the image element map Ma on which the one or more candidate points are superimposed to the display unit 140.

[0354] (Step S134) The display unit 140 acquires the image from the image processing unit 334 and displays the image, that is, the image including the image element map Ma on which one or more candidate points are superimposed.

[0355] By executing the processes of steps S111, S112, S121, and S131 to S134, the predicted property values ​​of the compound and the properties of the candidate points are displayed.

[0356] (Variation of Embodiment 2A) The image processing unit 334 may generate an image map Mb as in Variation 1 of Embodiment 1A and superimpose candidate points on the image map Mb. Note that the image map Mb is a map consisting of an arrangement of multiple maps Ma.

[0357] FIG. 34 is a diagram showing an example of the image map Mb in this modified example.

[0358] The image processing unit 334 superimposes candidate points on the image map Mb in accordance with the third display method information, as shown in FIG. 34 . Here, when the candidate point determination unit 331 acquires the third display method information indicating the limit number of candidate points from the input unit 110 via the display method acquisition unit 133, the candidate point determination unit 331 determines the limit number of candidate points. For example, the limit number is n (n is an integer equal to or greater than 1, e.g., 10) for the image map Mb. As a result, the image processing unit 334 superimposes, for example, 10 candidate points on the image map Mb.

[0359] This means that, for example, if the number of experiments that can be conducted in a week is 10, a user can easily grasp the candidate points that will be the subject of experiments in that week by setting the limit number n to n = 10.

[0360] FIG. 35 is a diagram showing another example of the image map Mb in this modified example.

[0361] 35, the candidate point determiner 331 may classify, in accordance with the third display method information, a plurality of candidate points superimposed on the image map Mb into candidate points whose predicted characteristic values ​​are equal to or greater than a reference value and candidate points whose predicted characteristic values ​​are less than the reference value, as in the example shown in Fig. 31. Note that the reference value used in the image map Mb may be, for example, the average value of the predicted characteristic values ​​at each position in the image map Mb, the median value of the predicted characteristic values, or an arbitrary value, but is not limited to these.

[0362] Specifically, for each of the multiple candidate points superimposed on the image map Mb, the candidate point determiner 331 determines whether the predicted characteristic value indicated at that candidate point is equal to or greater than a reference value, and if it determines that the predicted characteristic value is equal to or greater than the reference value, it attaches a flag indicating, for example, 1 to that candidate point. On the other hand, if it determines that the predicted characteristic value indicated at a candidate point is less than the reference value, it attaches a flag indicating, for example, 0 to that candidate point. Then, the candidate point determiner 331 outputs the multiple candidate points and the flags to the image processor 334.

[0363] 31 , when the image processing unit 334 acquires a plurality of candidate points and flags from the candidate point determination unit 331, it assigns, for example, a star-shaped mark to candidate points that have been assigned a flag indicating 1, as shown in FIG. 35 . The image processing unit 334 also assigns, for example, an inverted triangle mark to candidate points that have been assigned a flag indicating 0. The image processing unit 334 then superimposes the mark of the form assigned to the candidate point on the image map Mb, and displays an image including the image map Mb with the superimposed mark on the display unit 140. As a result, the candidate points indicated by the marks of each form are displayed superimposed on the image map Mb.

[0364] Alternatively, the candidate point determination unit 331 may use multiple reference values ​​to classify the multiple candidate points into three or more groups based on the comparison results between the reference values ​​and the predicted characteristic values ​​of each candidate point. Alternatively, as in the example of Figure 32, the candidate point determination unit 331 may use the above-mentioned priority instead of the predicted characteristic values ​​of the candidate points to classify the multiple candidate points into multiple groups based on the comparison results between the priority reference value and the priority of each candidate point. After classifying the multiple candidate points into two groups, as in Figure 35, for example, the candidate point determination unit 331 associates the multiple candidate points with group identification information indicating the groups to which the candidate points belong, and outputs the association information to the image processing unit 334.

[0365] When the image processing unit 334 acquires a plurality of candidate points and group identification information from the candidate point determination unit 331, the image processing unit 334 assigns, for example, a star-shaped mark to candidate points corresponding to group identification information indicating groups whose priorities are equal to or greater than the priority reference value, as shown in Fig. 35. The image processing unit 334 also assigns, for example, an inverted triangle mark to candidate points corresponding to group identification information indicating groups whose priorities are less than the priority reference value.

[0366] Then, the image processing unit 334 superimposes marks of the forms assigned to the candidate points on the image map Mb, and displays an image including the image map Mb with the superimposed marks on the display unit 140. As a result, the candidate points indicated by the marks of each form are displayed superimposed on the image map Mb.

[0367] This allows the user to easily understand the group to which each candidate point superimposed on the image map Mb belongs, i.e., the characteristics of the candidate points corresponding to that group, such as whether their characteristic prediction values ​​or priorities are high or low.

[0368] FIG. 36 is a diagram showing another example of the image map Mb in this modified example.

[0369] The candidate point determination unit 331 may set the number of candidate points to be superimposed on each image element map Ma included in the image map Mb to n (n is an integer greater than or equal to 1) in accordance with the third display method information. In the example shown in FIG. 36 , n=1. For example, the input unit 110 outputs third display method information indicating the number of candidate points to the display method acquisition unit 133 in response to a user's input operation on the input unit 110. The display method acquisition unit 133 acquires the third display method information and outputs the third display method information to the candidate point determination unit 331. The candidate point determination unit 331 receives the third display method information from the display method acquisition unit 133 and determines the number of candidate points indicated by the third display method information for each image element map Ma of the image map Mb. This allows the user to easily check the candidate points widely dispersed across the image map Mb.

[0370] Thus, in the characteristic display device 330 of this modified example, the image map Mb includes a plurality of image element maps Ma arranged in a matrix along the first and second coordinate axes, and each of the plurality of image element maps Ma has a third and fourth coordinate axis. The image processing unit 334 associates the first, second, third, and fourth coordinate axes with the first, second, third, and fourth variables used to represent the compound's structure, respectively. Then, for each of the plurality of compounds, the image processing unit 334 identifies an image element map Ma among the plurality of image element maps Ma associated with the number of the first variable and the value of the second variable used to represent the compound's structure. Next, the image processing unit 334 maps the predicted characteristic value of the compound to a position on the identified image element map Ma corresponding to the value of the third and fourth variables used to represent the compound's structure. For example, the first, second, third, and fourth variables are the variables a, b, x, and y described above.

[0371] This allows maps such as image map Mb to show predicted property values ​​for multiple compound structures, each represented by four variables, and to clearly display predicted property values ​​for compounds over a wide range, thereby improving the efficiency of materials discovery.

[0372] In the characteristic display device 330 of this modified example, the third display method information indicates the number of candidate points to be superimposed on each of the plurality of image element maps Ma as a display method of the candidate points. The image processing unit 334 superimposes the number of candidate points indicated by the third display method information on each of the plurality of image element maps Ma.

[0373] This allows the candidate points to be superimposed evenly on the multiple image element maps Ma included in the image map Mb, allowing the user to easily check the candidate points that are widely dispersed in the image map Mb.

[0374] (Embodiment 2B) Similar to embodiment 2A, the display system of this embodiment generates a map showing predicted property values ​​for each of a plurality of compounds, superimposes candidate points on the map, and further superimposes experimental property values ​​for each of one or more compounds on the map. Note that, among the components of this embodiment, the same components as those of embodiment 2A are assigned the same reference numerals as those of embodiment 2A, and detailed description thereof will be omitted. Furthermore, the map in this embodiment is an image element map Ma, but may also be an image map Mb.

[0375] [Configuration of Display System 400] Fig. 37 is a block diagram showing an example of the configuration of a display system 400 according to this embodiment. The display system 400 shown in Fig. 37 includes an input unit 110, a predictor database 120, a characteristic display device 430, a display unit 140, and an experiment database 150. The characteristic display device 430 is an example of an information display device.

[0376] The characteristic display device 430 in this embodiment superimposes not only the candidate points but also experimental characteristic values ​​of the compound obtained through experiments on the image element map Ma, and displays an image including the image element map Ma on which the candidate points and experimental characteristic values ​​are superimposed on the display unit 140. Such a characteristic display device 430 includes a search range acquisition unit 131, a predicted value acquisition unit 132, a display method acquisition unit 133, an experimental value acquisition unit 232, a candidate point determination unit 331, and an image processing unit 434. The characteristic display device 430 may be composed of a processor such as a CPU and a memory. In this case, the processor functions as the characteristic display device 430 by, for example, executing a computer program stored in the memory. The memory may be volatile or nonvolatile, or may be composed of a volatile memory and a nonvolatile memory.

[0377] [Experimental Value Acquisition Unit 232] The experimental value acquisition unit 232 in this embodiment acquires the structure and characteristic experimental values ​​of each of one or more compounds that have been tested from the experiment database 150. Then, the experimental value acquisition unit 232 outputs the acquired structure and characteristic experimental values ​​for each of the one or more compounds that have been tested to the candidate point determination unit 331 and also to the image processing unit 434.

[0378] [Image Processing Unit 434] The image processing unit 434 generates an image element map Ma, similar to embodiment 2A. Also, similar to embodiment 2A, the image processing unit 434 acquires third display method information from the display method acquisition unit 133 and superimposes one or more candidate points determined by the candidate point determination unit 331 on the image element map Ma according to the third display method information. Furthermore, the image processing unit 434 acquires the respective structures and characteristic experimental values ​​of one or more experimentally tested compounds from the experimental value acquisition unit 232. The image processing unit 434 then superimposes the respective characteristic experimental values ​​of the one or more experimentally tested compounds on the image element map Ma. In superimposing the characteristic experimental values ​​of the compounds, the image processing unit 434 superimposes the characteristic experimental values ​​of the compounds at positions on the image element map Ma corresponding to the structures of the compounds. That is, the image processing unit 434 superimposes marks indicating the characteristic experimental values ​​on the image element map Ma. Then, the image processing unit 434 outputs to the display unit 140 an image including the image element map Ma on which the experimental characteristic values ​​of one or more compounds that have been tested and one or more candidate points are superimposed.

[0379] [Specific Example of Map] FIG. 38 is a diagram showing an example of an image element map Ma on which candidate points and characteristic experimental values ​​are superimposed.

[0380] The image processing unit 434 superimposes the candidate points as marks such as star shapes on the image element map Ma in accordance with the third display method information. That is, the candidate point determination unit 331 determines one or more candidate points from all positions on the image element map Ma, and the image processing unit 434 superimposes star shapes on the one or more candidate points. Furthermore, the image processing unit 434 superimposes the experimental characteristic values ​​of the compounds as marks such as circles on the image element map Ma. That is, when the image processing unit 434 acquires the composition and experimental characteristic values ​​of one or more experimental compounds from the experimental value acquisition unit 232, it superimposes a circular mark having a color with a shade corresponding to the experimental characteristic value of the compound at a position on the image element map Ma corresponding to the composition of the compound.

[0381] As described above, in the characteristic display device 430 of this embodiment, the experimental value acquisition unit 232 acquires the experimental characteristic values ​​of each of the one or more compounds that have been tested. Then, the image processing unit 434 superimposes the experimental characteristic values ​​of each of the one or more compounds that have been tested on positions on the image element map Ma that correspond to the configuration of the compound. The image processing unit 434 generates an image including the image element map Ma on which one or more candidate points and one or more experimental characteristic values ​​are superimposed.

[0382] This allows the user to easily compare the candidate points in the image element map Ma with the experimental characteristic values ​​and the positions of those values. As a result, the user can easily determine whether or not to conduct experiments on the compounds corresponding to the candidate points based on the experimental characteristic values ​​and the positions of those values. This makes it possible to improve the efficiency of material search and appropriately support material development.

[0383] FIG. 39 is a diagram showing another example of the image element map Ma on which candidate points and characteristic experimental values ​​are superimposed.

[0384] The candidate point determination unit 331 determines, as the final candidate point, a position that is at least a predetermined distance away from any of the positions of the characteristic experimental values ​​among the plurality of tentative candidate points described above in accordance with the third display method information. The predetermined distance is, for example, three times the step width (i.e., one step) of each of the continuous variables x and y. In other words, the predetermined distance can also be considered three steps. This predetermined distance is also referred to as the second minimum separation distance.

[0385] The image processing unit 434 superimposes, on the image element map Ma, one or more candidate points determined as described above by the candidate point determination unit 331 in accordance with the third display method information. Furthermore, the image processing unit 434 also superimposes the experimental characteristic values ​​of the compound that has been tested at positions on the image element map Ma that correspond to the structure of the compound.

[0386] This makes it possible to exclude candidate points from positions corresponding to the structure of compounds that have been tested (i.e., positions of experimental characteristic values) and their surroundings, thereby enabling an efficient search for candidate points.

[0387] In other words, in the characteristic display device 430 of this embodiment, the third display method information indicates, as a display method for the candidate points, a predetermined second minimum separation distance between the candidate points and the positions of the characteristic experimental values. The image processing unit 434 superimposes one or more candidate points on the image element map Ma at a distance of at least the second minimum separation distance from each of the positions of one or more characteristic experimental values ​​acquired by the experimental value acquisition unit 232.

[0388] This makes it possible to prevent candidate points from overlapping near the positions of experimental property values. As a result, it is possible to prevent experiments from being conducted on compounds (i.e., candidate compounds) that have structures similar to compounds already tested that have experimental property values, thereby preventing the repeated acquisition of similar experimental results and increasing the possibility of obtaining novel experimental results. As a result, it is possible to improve the efficiency of materials search.

[0389] FIG. 40 is a diagram showing an example of the state transition of the image element map Ma.

[0390] For example, when the experimental value acquisition unit 232 has not acquired the characteristic experimental value, the image processing unit 434 displays the image element map Ma shown in Fig. 40(a) on the display unit 140. A plurality of candidate points are superimposed on the image element map Ma shown in Fig. 40(a), but the characteristic experimental value is not superimposed.

[0391] Here, the user conducts an experiment on one candidate point b1 of the multiple candidate points in the image element map Ma. Specifically, the user conducts an experiment on a compound having the structure indicated by candidate point b1. This experiment updates the experimental data in the experimental database 150. That is, the structure and experimental characteristic values ​​of the new compound are written to the experimental data. As a result, the experimental value acquisition unit 232 acquires the structure and experimental characteristic values ​​of the new compound and outputs them to the candidate point determination unit 331 and the image processing unit 434. Upon acquiring the structure and experimental characteristic values ​​of the new compound from the experimental value acquisition unit 232, the image processing unit 434 superimposes the experimental characteristic values ​​at a position on the image element map Ma corresponding to the structure, as shown in (b) of FIG. 40 . That is, the image element map Ma with the new experimental characteristic values ​​superimposed at the position of candidate point b1 is displayed on the display unit 140.

[0392] Next, the user operates the input unit 110 so that only candidate points that are at least the second minimum distance from the position of the characteristic experimental value are displayed. This causes the display method acquisition unit 133 to acquire information from the input unit 110 in response to the user's input operation. That is, the display method acquisition unit 133 acquires third display method information indicating a display method in which only candidate points that are at least the second minimum distance from the position of the characteristic experimental value are displayed. The display method acquisition unit 133 then outputs the third display method information to the candidate point determination unit 331 and the image processing unit 434. As a result, the candidate point determination unit 331, as in the example of FIG. 39 , retains only candidate points that are at least the second minimum distance from the position of the characteristic experimental value among the multiple candidate points superimposed on the image element map Ma shown in FIG. 40 (b), and deletes candidate points that are not at least the second minimum distance from the position of the characteristic experimental value. That is, the candidate point determination unit 331 re-determines and updates the candidate points. The image processing unit 434 displays an image element map Ma on the display unit 140, in which one or more re-determined candidate points and characteristic experimental values ​​are superimposed, thereby enabling the user to proceed with the experiment more efficiently.

[0393] In the above example, the image element map Ma shown in FIG. 40(b) is displayed on the display unit 140. However, if the third display method information has already been acquired before the characteristic experimental value is acquired, the image element map Ma shown in FIG. 40(b) does not need to be displayed. That is, the image element map Ma shown in FIG. 40(a) may be displayed, and after the experiment is conducted, the image element map Ma shown in FIG. 40(c) may be directly displayed. Furthermore, after the image element map Ma shown in FIG. 40(c) is displayed, if the experiment is further conducted and a new characteristic experimental value is obtained, the same processing as described above is repeatedly executed. That is, among the multiple candidate points superimposed on the image element map Ma shown in FIG. 40(c), only those candidate points that are at least the second minimum distance from the position of the new characteristic experimental value are retained, and candidate points that are not at least the second minimum distance from the position of the new characteristic experimental value are deleted. As a result, the multiple candidate points superimposed on the image element map Ma shown in FIG. 40(c) are updated.

[0394] FIG. 41 is a diagram showing an example of the state transition of the image element map Ma.

[0395] As in the example of Fig. 30 , the candidate point determiner 331 determines a plurality of candidate points and an experimental design, as shown in Fig. 41(a). The experimental design is information indicating which of the plurality of candidate points should be tested and in what order. The candidate point determiner 331 outputs information indicating the determined plurality of candidate points and the experimental design to the image processor 434.

[0396] The image processing unit 434 acquires information indicating a plurality of candidate points and an experimental design from the candidate point determination unit 331. The image processing unit 434 then superimposes the plurality of candidate points on the image element map Ma. Furthermore, the image processing unit 434 assigns the experiment numbers of the n design candidate points included in the experimental design to the image element map Ma, and further superimposes arrows pointing from one design candidate point to the design candidate point of the next experiment number on the image element map Ma.

[0397] Here, the user conducts an experiment on a compound (i.e., a candidate compound) having a structure indicated by the design candidate point with experiment number (1), for example. This results in the acquisition of experimental characteristic values ​​for the compound, and the experimental data in the experimental database 150 is updated. That is, the structure and experimental characteristic values ​​of the new compound obtained through the experiment are written to the experimental data. As a result, the experimental value acquisition unit 232 acquires the structure and experimental characteristic values ​​of the new compound and outputs them to the candidate point determination unit 331 and the image processing unit 434. Upon acquiring the structure and experimental characteristic values ​​of the new compound from the experimental value acquisition unit 232, the image processing unit 434 superimposes the experimental characteristic values ​​on the position on the image element map Ma corresponding to the structure, i.e., the position of the design candidate point with experiment number (1), as shown in FIG. 41B. That is, the new experimental characteristic values ​​are superimposed on the image element map Ma in place of the design candidate point with experiment number (1).

[0398] Furthermore, the candidate point determiner 331 acquires the structure and experimental characteristic values ​​of the new compound from the experimental value acquirer 232. At this time, the candidate point determiner 331, as in the example of FIG. 40 , retains only candidate points that are separated from the position of the experimental characteristic values ​​by a second minimum distance or more from the multiple candidate points superimposed on the image element map Ma shown in FIG. 41 (a). The candidate point determiner 331 then deletes candidate points that are not separated from the position of the experimental characteristic values ​​by a second minimum distance or more. In other words, by re-determining and updating the candidate points, the candidate point determiner 331 deletes, for example, three candidate points that are not separated by a second minimum distance or more from the position of the experimental characteristic values ​​on which the planned candidate points for experiment number (1) were superimposed, as shown in FIG. 41 (b). The image processing unit 434 displays on the display unit 140 an image element map Ma in which one or more candidate points thus re-determined are superimposed on characteristic experimental values, i.e., an image element map Ma in which multiple candidate points including the four planned candidate points with experiment numbers (2) to (5) are superimposed on characteristic experimental values.

[0399] Here, the user further conducts an experiment on a compound (i.e., a candidate compound) having a structure indicated by the design candidate point of experiment number (2), for example. This obtains experimental characteristic values ​​for the compound, and the experimental data in the experiment database 150 is updated. As a result, as described above, the experimental value acquisition unit 232 acquires the structure and experimental characteristic values ​​of the new compound and outputs them to the candidate point determination unit 331 and the image processing unit 434. Upon acquiring the structure and experimental characteristic values ​​of the new compound from the experimental value acquisition unit 232, the image processing unit 434 superimposes the experimental characteristic values ​​on the position on the image element map Ma corresponding to the structure, i.e., the position of the design candidate point of experiment number (2), as shown in (c) of FIG. 41 . In other words, the new experimental characteristic values ​​are superimposed on the image element map Ma in place of the design candidate point of experiment number (2).

[0400] Furthermore, when experimental data is updated, the predictors in the predictor database 120 may be updated to the latest state in accordance with the updated experimental data. For example, if the predictor is a model obtained by supervised machine learning, the predictor is retrained using the structures and experimental property values ​​of compounds newly added to the experimental data as training data. As a result, the predicted value acquisition unit 132 acquires the updated predictor from the predictor database 120 and uses the updated predictor to reacquire predicted property values ​​corresponding to the structures of the multiple compounds. The candidate point determination unit 331 acquires the structures and predicted property values ​​of the multiple compounds from the predicted value acquisition unit 132 and re-determines one or more candidate points based on the structures and predicted property values. As a result, for example, as shown in (c) of FIG. 41, two new candidate points are added around the design candidate point for experiment number (5).

[0401] 41(c), the candidate point determiner 331 deletes, as shown in FIG. 41(b), for example, two candidate points that are not more than the second minimum separation distance from the position of the characteristic experimental value on which the design candidate point for experiment number (2) was superimposed. The image processor 434 displays on the display unit 140 an image element map Ma on which one or more re-determined candidate points and the characteristic experimental value are superimposed, i.e., an image element map Ma on which multiple candidate points including the three design candidate points for experiment numbers (3) to (5) and two characteristic experimental values ​​are superimposed. This allows the user to proceed with the experiment more efficiently.

[0402] In this way, in the characteristic display device 430 of this embodiment, after the image element map Ma is displayed, the processing by the experimental value acquisition unit 232 is further repeated, and when a new characteristic experimental value is acquired, the image processing unit 434 updates one or more candidate points that are already superimposed on the image element map Ma so that all candidate points superimposed on the image element map Ma are separated from the new characteristic experimental value by at least the second minimum separation distance.

[0403] This allows new experimental property values ​​to be acquired each time an experiment on a candidate compound is conducted, and appropriate candidate points can be displayed according to the new experimental property values, thereby improving the efficiency of materials search and providing appropriate support for materials development.

[0404] [Processing Operation] FIG. 42 is a flowchart showing the processing operation of the display system 400 in this embodiment.

[0405] Similar to the display system 300 of embodiment 2A, the display system 400 of this embodiment executes the processes of steps S111, S112, S121, S131, and S132 shown in Fig. 33. Next, the display system 400 executes the processes of steps S141 and S142.

[0406] (Step S141) The image processing unit 434 acquires the structure and predicted property values ​​of each of the multiple compounds from the predicted value acquisition unit 132. Furthermore, the image processing unit 434 acquires one or more candidate points from the candidate point determination unit 331. Furthermore, the image processing unit 434 acquires the structure and experimental property values ​​of each of the one or more compounds that have been tested from the experimental value acquisition unit 232. Furthermore, the image processing unit 434 receives third display method information from the display method acquisition unit 133.

[0407] The image processing unit 434 generates an image element map Ma based on the acquired structures and predicted property values ​​of each of the multiple compounds. Furthermore, the image processing unit 434 superimposes one or more candidate points acquired from the candidate point determination unit 331 on the image element map Ma, and superimposes one or more experimental property values ​​acquired from the experimental value acquisition unit 232 on the image element map Ma. The superimposition of the candidate points is performed in accordance with the third display method information. The image processing unit 434 outputs an image including the image element map Ma on which the one or more candidate points and experimental property values ​​are superimposed to the display unit 140.

[0408] (Step S142) The display unit 140 acquires an image from the image processing unit 434 and displays the image, that is, an image including the image element map Ma on which one or more candidate points and characteristic experimental values ​​are superimposed.

[0409] By executing the processes of steps S111, S112, S121, S131, S132, S141, and S142, the predicted property values, candidate points, and experimental property values ​​of the compound are displayed, allowing the user to properly recognize the overall picture of the experimental property values ​​on the image element map Ma while comparing them with the predicted property values ​​and candidate points.

[0410] (Variation of Embodiment 2B) The display system in this variation has a function of saving and displaying an image generated by the characteristic display device 430, i.e., an image including an image element map Ma on which one or more candidate points and characteristic experimental values ​​are superimposed. Note that, among the components in this variation, the same components as those in embodiment 2B are assigned the same reference numerals as those in embodiment 2B, and detailed description thereof will be omitted.

[0411] [Configuration of display system 401] Fig. 43 is a block diagram showing an example of the configuration of display system 401 in this modification. Display system 401 shown in Fig. 43 includes all of the components included in display system 400 in embodiment 2B, and an image storage unit 410.

[0412] [Image Storage Unit 410] The image storage unit 410 is a recording medium for storing, as a characteristic display image, an image including an image element map Ma on which one or more candidate points and characteristic experimental values ​​are superimposed, generated by the image processing unit 434. This recording medium is, for example, a hard disk drive, RAM, ROM, or semiconductor memory. The recording medium may be volatile or non-volatile.

[0413] [Image Processing Unit 434] As in embodiment 2B, the image processing unit 434 in this modification generates a characteristics display image and stores the generated characteristics display image in the image storage unit 410. At this time, the image processing unit 434 may add date information indicating the date on which the characteristics display image was generated to the characteristics display image, or may add reference information such as the search range used to generate the characteristics display image.

[0414] Furthermore, when the image processing unit 434 receives third display method information indicating a display method for a past characteristic display image from the display method acquisition unit 133, it reads out one or more characteristic display images from the image storage unit 410 in accordance with the third display method information. Then, the image processing unit 434 generates a history list image including the one or more characteristic display images and outputs it to the display unit 140.

[0415] [Specific Example of History List Image] FIG. 44 is a diagram showing an example of the history of the characteristic display image.

[0416] The input unit 110 accepts an input operation by a user and outputs third display method information indicating a search range and period corresponding to the input operation as a display method for past characteristic display images to the display method acquisition unit 133. The period is, for example, January to March 2021. The display method acquisition unit 133 acquires the third display method information and outputs the third display method information to the image processing unit 434. Upon acquiring the third display method information from the display method acquisition unit 133, the image processing unit 434 searches the image storage unit 410 for one or more characteristic display images corresponding to the third display method information. In other words, the image processing unit 434 searches the image storage unit 410 for characteristic display images to which reference information indicating the same search range as the search range indicated by the third display method information is added and to which date information indicating a date within the period indicated by the third display method information is added.

[0417] Then, the image processing unit 434 generates a history list image by arranging one or more characteristic display images found by the search in the order of the dates of the date information added to them, and outputs the history list image to the display unit 140. As a result, the history list image is displayed on the display unit 140, for example, as shown in Fig. 44.

[0418] For example, the history list image includes a characteristic display image generated on January 20, 2021 shown in (a) of Fig. 44, a characteristic display image generated on February 20, 2021 shown in (b) of Fig. 44, and a characteristic display image generated on March 20, 2021 shown in (c) of Fig. 44. These characteristic display images are arranged in chronological order from oldest to newest.

[0419] This allows the user to easily grasp the update status of the characteristic experiment values ​​and candidate points for each month during the period designated by the user's input operation, i.e., three months.

[0420] In the example of FIG. 44 , the history list image includes three characteristic display images. However, if there are four or more characteristic display images corresponding to the third display method information, a history list image including four or more characteristic display images may be displayed. Conversely, if there are one or two characteristic display images corresponding to the third display method information, a history list image including one or two characteristic display images may be displayed. Furthermore, if there are a predetermined number (e.g., 100) or more characteristic display images corresponding to the third display method information, those characteristic display images may be thinned out. As a result, a history list image including fewer than the predetermined number of characteristic display images is displayed. This prevents the history list image from including too many characteristic display images, making each characteristic display image easier to see.

[0421] In this way, in the characteristic display device 430 of this embodiment, after the generation and display of an image including the above-mentioned image element map Ma are repeatedly executed, the image processing unit 434 reads out the plurality of images from the image storage unit 410, generates a composite image including the plurality of images, and outputs it to the display unit 140. This composite image is the above-mentioned history list image. Note that in order to repeatedly execute the generation and display of the above-mentioned image, the acquisition of predicted characteristic values, generation of the image element map Ma, acquisition of third display method information, acquisition of experimental characteristic values, superimposition of candidate points, superimposition of experimental characteristic values, and image generation are repeatedly executed.

[0422] This allows the user to review the progress of the experiment so far and appropriately determine the direction of future material searches, thereby improving the efficiency of material searches.

[0423] (Embodiment 2C) Similar to embodiment 2B, the display system of this embodiment generates a map showing predicted property values ​​for each of a plurality of compounds and superimposes candidate points and experimental property values ​​on the map. Here, the property display device of this embodiment acquires the candidate points from a database, without determining them from predicted property values ​​or the like. Note that, among the components of this embodiment, the same components as those of embodiments 2A and 2B are assigned the same reference numerals as those of embodiments 2A and 2B, and detailed description thereof will be omitted. Furthermore, the map in this embodiment is an image element map Ma, but may also be an image map Mb.

[0424] [Configuration of Display System 500] Fig. 45 is a block diagram showing an example of the configuration of a display system 500 according to this embodiment. The display system 500 shown in Fig. 45 includes an input unit 110, a predictor database 120, a characteristic display device 530, a display unit 140, an experiment database 150, and a candidate point database (DB) 510. The characteristic display device 530 is an example of an information display device.

[0425] The characteristic display device 530 in this embodiment includes a search range acquisition unit 131, a predicted value acquisition unit 132, a display method acquisition unit 133, an experimental value acquisition unit 232, a candidate point acquisition unit 531, and an image processing unit 534. The characteristic display device 530 may be composed of a processor such as a CPU and a memory. In this case, the processor functions as the characteristic display device 530 by executing a computer program stored in the memory, for example. The memory may be volatile or non-volatile, or may be composed of a volatile memory and a non-volatile memory.

[0426] The predicted value acquisition unit 132 in this embodiment outputs the structure and characteristic predicted values ​​of each of the multiple compounds to the image processing unit 534. The experimental value acquisition unit 232 in this embodiment outputs the structure and characteristic experimental values ​​of each of the one or more compounds to the image processing unit 534. Furthermore, the display method acquisition unit 133 in this embodiment outputs third display method information to the image processing unit 534. That is, the predicted value acquisition unit 132, the experimental value acquisition unit 232, and the display method acquisition unit 133 in this embodiment do not output information for determining candidate points to the candidate point acquisition unit 531.

[0427] [Candidate Point Acquisition Unit 531] The candidate point acquisition unit 531 acquires, for example, a search range signal from the search range acquisition unit 131. Then, the candidate point acquisition unit 531 searches the candidate point database 510 for one or more compounds having a structure included in the search range indicated by the search range signal. The candidate point acquisition unit 531 acquires the structure of the compound found by the search as a candidate point from the candidate point database 510 and outputs the candidate point to the image processing unit 534.

[0428] [Image Processing Unit 534] The image processing unit 534 acquires the predicted structure and predicted property values ​​of each of the multiple compounds from the predicted value acquisition unit 132, and acquires the experimental structure and experimental property values ​​of each of the one or more compounds that have been tested from the experimental value acquisition unit 232. The image processing unit 534 also acquires one or more candidate points from the candidate point acquisition unit 531 and receives third display method information from the display method acquisition unit 133. The image processing unit 534 generates an image element map Ma showing the predicted property values ​​of the multiple compounds and superimposes the experimental property values ​​of each of the one or more compounds that have been tested on the image element map Ma. Furthermore, the image processing unit 534 superimposes one or more candidate points on the image element map Ma in accordance with the third display method information. The image processing unit 534 then outputs an image including the image element map Ma on which the experimental property values ​​of the one or more compounds and one or more candidate points are superimposed to the display unit 140.

[0429] [Candidate Point Database 510] The candidate point database 510 stores candidate point data indicating the respective structures of a plurality of compounds.

[0430] FIG. 46 is a diagram showing an example of candidate point data stored in the candidate point database 510. For example, as shown in FIG. 46, the candidate point data indicates, for each of a plurality of compounds, the compound's control number, composition formula, firing method, firing temperature, and firing time as the compound's configuration. The firing method, firing temperature, and firing time are process conditions for firing the compound. For example, as shown in FIG. 46, the candidate point data indicates, for one compound, the control number "1," the composition formula "Li1.45La0.90Ti1.0Ga0.10O3," the firing method "1: solid-phase method," the firing temperature "100°C," and the firing time "3 hours."

[0431] The candidate point acquisition unit 531, for example, acquires a search range signal from the search range acquisition unit 131 and acquires, as candidate points, composition formulas of compounds included in the search range indicated by the search range signal from the candidate point data. That is, the candidate point acquisition unit 531 generates all combinations of option data for each of the variables M3, M3', M4, M4', x, y, a, and b included in the search range. Then, for each combination, the candidate point acquisition unit 531 acquires, as candidate points, composition formulas having that combination from the candidate point data. Furthermore, if the search range includes process condition variables Pa, Pb, and Pc, the candidate point acquisition unit 531 generates all combinations of option data for each of the multiple variables including the variables Pa, Pb, and Pc. Then, for each combination, the candidate point acquisition unit 531 acquires, as candidate points, composition formulas and process conditions corresponding to that combination from the candidate point data. Alternatively, the candidate point acquisition unit 531 may acquire, as candidate points, the composition formulas or the composition formulas and process conditions of all compounds indicated in the candidate point data, regardless of the search range.

[0432] [Specific Example of Map] FIG. 47 is a diagram showing an example of an image element map Ma on which candidate points and characteristic experimental values ​​are superimposed.

[0433] The candidate point acquisition unit 531 acquires each structure of one or more compounds as a candidate point from the candidate point database 510. At this time, the candidate point acquisition unit 531 acquires not only the candidate points but also the control numbers associated with the structures (i.e., composition formulas, etc.) that are the candidate points. The candidate point acquisition unit 531 then outputs the candidate points and the control numbers to the image processing unit 534.

[0434] The image processing unit 534 generates an image element map Ma showing predicted property values ​​for each of the multiple compounds, as shown in FIG. 47 . Then, the image processing unit 534 superimposes one or more candidate points acquired by the candidate point acquisition unit 531 and experimental property values ​​for each of the one or more tested compounds acquired by the experimental value acquisition unit 232 on the image element map Ma. Furthermore, the image processing unit 534 superimposes a reference image a1 showing each of the one or more candidate points and the control number acquired by the candidate point acquisition unit 531 on the image element map Ma in association with the candidate points. The reference image may have the form of a speech bubble extending from the candidate point. Then, the image processing unit 534 outputs an image including the image element map Ma on which the experimental property values ​​for the one or more tested compounds, one or more candidate points, and the reference image a1 are superimposed to the display unit 140. As a result, the image element map Ma is displayed on the display unit 140, as shown in FIG. 47 .

[0435] Here, the image processing unit 534 may display the process conditions on the display unit 140. For example, the user selects one candidate point by performing an input operation on the input unit 110. More specifically, the user uses a mouse or the like included in the input unit 110 to position a cursor on a desired candidate point and selects the candidate point by clicking the mouse button. As a result, the display method acquisition unit 133 acquires an input signal corresponding to the input operation from the input unit 110 and notifies the image processing unit 534 of the candidate point selected by the user, which is indicated by the input signal. Upon receiving the notification, the image processing unit 534 displays a reference image indicating the process conditions corresponding to the selected candidate point on the display unit 140. For example, the process conditions may be displayed as part of the reference image a1 described above.

[0436] Alternatively, the image processing unit 534 may superimpose in advance on the image element map Ma a reference image indicating the control number and process conditions for each of one or more candidate points, regardless of the user's input operation.

[0437] Alternatively, the image processing unit 534 may first display an image element map Ma not including the reference image a1 on the display unit 140, and after the user performs an input operation on the input unit 110, display the image element map Ma shown in FIG. 47 on the display unit 140. That is, after the user performs an input operation, a reference image a1 indicating a control number for each candidate point is superimposed on the image element map Ma, and an image including the image element map Ma is displayed on the display unit 140. Alternatively, if the user's input operation is an operation to select one candidate point, the image processing unit 534 may superimpose a reference image a1 only for the selected candidate point on the image element map Ma, and display an image including the image element map Ma on the display unit 140.

[0438] As described above, in the characteristic display device 530 of this embodiment, the candidate point acquisition unit 531 acquires one or more candidate points from the candidate point database 510, and the image processing unit 534 superimposes the acquired one or more candidate points on the image element map Ma. The candidate point database 510 stores candidate point data indicating the structure of a compound. For example, as shown in FIG. 46 , the candidate point database 510 stores candidate point data indicating a plurality of candidate points (i.e., a plurality of composition formulas).

[0439] As a result, since the candidate points are acquired from a database such as the candidate point database 510, the candidate points can be superimposed on the image element map Ma regardless of the predicted property values ​​of each of the multiple compounds acquired by the predicted value acquisition unit 132. Therefore, candidate points that are considered important in advance can be displayed on the image element map Ma. As a result, the efficiency of material search can be improved.

[0440] [Processing Operation] FIG. 48 is a flowchart showing the processing operation of the display system 500 in this embodiment.

[0441] Similar to display system 400 of embodiment 2B, display system 500 of the present embodiment executes the processes of steps S111, S112, S121, and S131 shown in Fig. 42. Furthermore, display system 500 executes the processes of steps S151, S152, and S153.

[0442] (Step S151) The candidate point acquisition unit 531 acquires each of one or more candidate points and the management number associated with the candidate point from the candidate point data in the candidate point database 510. Then, the candidate point acquisition unit 531 outputs the acquired one or more candidate points and their management numbers to the image processing unit 534.

[0443] (Step S152) The image processing unit 534 acquires the structure and predicted property values ​​of each of the multiple compounds from the predicted value acquisition unit 132. Furthermore, the image processing unit 534 acquires one or more candidate points and their management numbers from the candidate point acquisition unit 531. Furthermore, the image processing unit 534 acquires the structure and experimental property values ​​of each of the one or more compounds that have been tested from the experimental value acquisition unit 232. The image processing unit 534 also receives third display method information from the display method acquisition unit 133.

[0444] The image processing unit 534 generates an image element map Ma based on the structures and predicted property values ​​of each of the acquired compounds. Furthermore, the image processing unit 534 superimposes one or more candidate points acquired from the candidate point acquisition unit 531 on the image element map Ma, and superimposes reference images a1 indicating the control numbers corresponding to the candidate points on the image element map Ma in association with the candidate points. The superimposition of the candidate points is performed in accordance with the third display method information. Furthermore, the image processing unit 534 superimposes one or more experimental property values ​​acquired from the experimental value acquisition unit 232 on the image element map Ma. The image processing unit 534 outputs an image including the image element map Ma on which the one or more candidate points, the reference image a1, and the experimental property values ​​are superimposed to the display unit 140.

[0445] (Step S153) The display unit 140 acquires an image from the image processing unit 534 and displays the image, that is, the image element map Ma on which one or more candidate points, the reference image a1, and the characteristic experimental value are superimposed.

[0446] By executing the processes of steps S111, S112, S121, S131, and S151 to S153, the predicted property values, candidate points, control numbers, and experimental property values ​​of the compound are displayed. As a result, the user can properly recognize the overall picture of the predicted property values, experimental property values, and candidate points on the image element map Ma, and can also easily grasp the control numbers associated with the candidate points.

[0447] (Embodiment 3A) A display system according to this embodiment displays the property evaluation values ​​of a plurality of compounds in the form of a map, instead of the predicted property values ​​of each compound used in Embodiments 1 and 2. The property evaluation values ​​are evaluation values ​​for the properties of compounds having structures corresponding to positions on the map and are treated as experimental indicators of the compound properties. These property evaluation values ​​indicate values ​​according to a selected calculation method. Therefore, if the calculation method is the same as the calculation method used to obtain the predicted property values ​​of Embodiments 1 and 2, the display system according to this embodiment displays the predicted property values ​​in the form of a map as the property evaluation values, similar to Embodiments 1 and 2. Furthermore, the display system according to this embodiment superimposes the respective structures of one or more compounds that have been tested as experimental points on the map. Furthermore, the display system according to this embodiment displays an image including the map as a first image, and when the display method for the property evaluation values ​​is changed, changes the first image to a second image.

[0448] Of the components in this embodiment, the same components as those in the first and second embodiments are denoted by the same reference numerals as those in the first and second embodiments, and detailed description thereof will be omitted.

[0449] [Configuration of Display System 600] Fig. 49 is a block diagram showing an example of the configuration of a display system 600 according to the present embodiment. The display system 600 shown in Fig. 49 includes an input unit 110, an evaluator database (DB) 620, a characteristic display device 630, a display unit 140, an evaluation display database (DB) 640, and an experiment database 650. The characteristic display device 630 is an example of an information display device.

[0450] The property display device 630 in this embodiment acquires property evaluation values ​​for each of a plurality of compounds using the evaluator database 620 and the evaluation display database 640. The property display device 630 then generates a map showing the acquired property evaluation values. Furthermore, the property display device 630 displays experimental points for each of one or more compounds superimposed on the map on the display unit 140. The property display device 630 includes an evaluation value acquisition unit 632, a display method acquisition unit 633, an experimental point acquisition unit 635, and an image processing unit 634. The property display device 630 may be composed of a processor, such as a CPU, and a memory. In this case, the processor functions as the property display device 630 by executing a computer program stored in the memory. The memory may be volatile or nonvolatile, or may be composed of a volatile memory and a nonvolatile memory. The map in this embodiment is an image map Mb, but it may also be an image element map Ma or a map composed of multiple image maps Mb.

[0451] [Display method acquisition unit 633] The display method acquisition unit 633 acquires an input signal from the input unit 110, and in response to the input signal, acquires first information d10 related to a method for displaying characteristic evaluation values ​​from the evaluation display database 640. The display method acquisition unit 633 then outputs the first information d10 to the evaluation value acquisition unit 632, the experimental point acquisition unit 635, and the image processing unit 634. As a result, an image including a map according to the display method indicated by the first information d10 is displayed on the display unit 140 as a first image.

[0452] Furthermore, when the display method acquisition unit 633 acquires an input signal indicating a change in the display method from the input unit 110, it changes the first information d10 in accordance with the input signal, and outputs the changed first information d10 to the evaluation value acquisition unit 632, the experiment point acquisition unit 635, and the image processing unit 634. As a result, an image including a map according to the changed first information d10 is displayed on the display unit 140 as the second image.

[0453] [Evaluation value acquisition unit 632] The evaluation value acquisition unit 632 acquires the first information d10 from the display method acquisition unit 633. Then, the evaluation value acquisition unit 632 generates all combinations of option data that can be taken by each of the multiple variables included in the search range indicated by the first information d10. The combinations represent the structure of the compound (e.g., composition formula). Therefore, multiple structures are generated according to these combinations.

[0454] The evaluation value acquisition unit 632 acquires one or more evaluators from the evaluator database 620 and inputs the generated configurations to each of the one or more evaluators. As a result, the evaluation value acquisition unit 632 acquires, for each configuration, a characteristic evaluation value for one or more characteristics of the compound having that configuration. At this time, the evaluation value acquisition unit 632 acquires those characteristic evaluation values ​​in accordance with the first information d10.

[0455] [Experimental Point Acquisition Unit 635] The experimental point acquisition unit 635 acquires the first information d10 from the display method acquisition unit 633. Then, the experimental point acquisition unit 635 acquires the respective configurations of one or more compounds that have been tested and that correspond to the search range indicated by the first information d10 from the experimental database 650, and outputs each configuration of the one or more compounds as an experimental point to the image processing unit 634. Note that the experimental point acquisition unit 635 may acquire the respective configurations of all compounds stored in the experimental database 650 as experimental points, regardless of the search range.

[0456] [Image Processing Unit 634] The image processing unit 634 acquires the characteristic evaluation values ​​of each of the multiple compounds from the evaluation value acquisition unit 632 and acquires the experimental points of each of the one or more compounds that have been tested from the experimental point acquisition unit 635. Furthermore, the image processing unit 634 receives first information d10 from the display method acquisition unit 633. The image processing unit 634 generates a map showing the characteristic evaluation values ​​of the multiple compounds according to the first information d10 and further superimposes one or more experimental points acquired by the experimental point acquisition unit 635 on the map. In generating the map showing the characteristic evaluation values, the image processing unit 634 applies a color with a shading corresponding to the characteristic evaluation value of each of the multiple compounds to a position on the map corresponding to the structure of the compound. The shading color corresponding to the characteristic evaluation value may be a color corresponding to the characteristic evaluation value. In superimposing the experimental points, the image processing unit 634 superimposes a black circular mark on a position on the map corresponding to the experimental point of each of the one or more compounds that have been tested. The image processing unit 634 outputs an image including a map on which one or more experimental points are superimposed to the display unit 140 as a first image or a second image.

[0457] [Evaluator Database 620] The evaluator database 620 is a recording medium that stores at least one evaluator for evaluating the property values ​​of compounds. This recording medium is, for example, a hard disk drive, RAM, ROM, or semiconductor memory. Furthermore, the recording medium may be volatile or non-volatile.

[0458] The evaluator is a computer program based on a predetermined calculation algorithm constructed by, for example, machine learning. In response to an input of a configuration (i.e., a combination of multiple variables) from the evaluation value acquisition unit 632, the evaluator outputs an evaluation value of the characteristic value of a compound having that configuration as a characteristic evaluation value. Note that the configuration is expressed by a combination of option data that can be taken by each of multiple variables included in the search range.

[0459] The property evaluation value is an index of the experiment, and is, for example, a predicted value of a property obtained by a machine learning model such as Least Absolute Shrinkage and Selection Operator (LASSO) regression or Random Forest regression. The predicted value corresponds to the property predicted value in the first and second embodiments. The property evaluation value may also be the output value of an acquisition function in Bayesian optimization. The acquisition function is a function for determining the next experimental candidate (i.e., a candidate compound configuration to be used in the next experiment, or a candidate compound in the second embodiment). Examples of the acquisition function include Expected Improvement (EI), Probability of Improvement (PI), and Upper Confidence Bound (UCB), which use the mean value and variance value of the predicted values ​​obtained by Gaussian Process regression.

[0460] Furthermore, one evaluator may output characteristic evaluation values ​​for various compound characteristics, such as electrochemical characteristics and thermochemical characteristics. The evaluator database 620 may store multiple evaluators, and these multiple evaluators may be used appropriately depending on the characteristics of the evaluation target. For example, the evaluator database 620 stores evaluator A, which outputs characteristic evaluation values ​​for electrochemical characteristics, and evaluator B, which outputs characteristic evaluation values ​​for thermochemical characteristics. In this case, when the characteristics of the evaluation target are, for example, electrochemical characteristics, the evaluation value acquisition unit 632 acquires evaluator A and uses evaluator A to acquire characteristic evaluation values ​​for the electrochemical characteristics of each of the multiple compounds. Alternatively, when the characteristics of the evaluation target are, for example, thermochemical characteristics, the evaluation value acquisition unit 632 acquires evaluator B and uses evaluator B to acquire characteristic evaluation values ​​for the thermochemical properties of each of the multiple compounds. This makes it possible to suppress the complexity of the calculation algorithm of each evaluator, thereby improving the evaluation accuracy and processing speed for each characteristic. It should be noted that the present disclosure is not limited to the above example, and the evaluator may be of any type as long as it can output a characteristic evaluation value for a compound having a configuration expressed by a combination of multiple variables. Evaluation of such characteristic values ​​can also be considered as searching for the characteristic values ​​of the compound.

[0461] [Evaluation Display Database 640] The evaluation display database 640 is a recording medium that stores information related to the display method of the characteristic evaluation value. This recording medium is, for example, a hard disk drive, RAM, ROM, or semiconductor memory. Furthermore, the recording medium may be volatile or non-volatile.

[0462] FIG. 50 is a diagram showing an example of information stored in the evaluation display database 640. As shown in FIG.

[0463] 50, the evaluation display database 640 stores first information d10, which is information relating to a display method of characteristic evaluation values. This first information d10 includes first color information d11, calculation method information d12, map arrangement information d13, search range information d14, first display target information d15, and display range information d16.

[0464] The first color information d11 indicates at least one of the hue, saturation, and brightness of the characteristic evaluation values ​​shown in the map as a color attribute.

[0465] The calculation method information d12 is information about a calculation method of the characteristic evaluation value, i.e., information applied to the evaluator to calculate the characteristic evaluation value. The calculation method information d12 may be, for example, information indicating the type of evaluator used to calculate the characteristic evaluation value, specifically the type of machine learning model, or may be parameters used in the evaluator. For example, if the evaluator is a neural network, the parameters may be weights, biases, or the like used in the neural network.

[0466] The map arrangement information d13 is information indicating the form of a map generated by the image processing unit 634. For example, if the map generated by the image processing unit 634 includes a plurality of image element maps Ma having coordinate axes A1 and A2 indicating variables x and y, like the image map Mb shown in Fig. 11, the map arrangement information d13 indicates the arrangement method of these image element maps Ma.

[0467] The search range information d14 indicates a search range used to calculate the characteristic evaluation value. The first display object information d15 indicates the characteristics of the display object related to the characteristic evaluation value. The display range information d16 indicates the display range of the characteristic evaluation value. Details of each of the above information included in the first information d10 will be described later.

[0468] [Experiment Database 650] FIG. 51 is a diagram showing an example of each data stored in the experiment database 650. As shown in FIG.

[0469] As shown in FIG. 51 , the experimental database 650 is a recording medium that stores compound basic data 651 and one or more compound detailed data 652 as experimental data. This recording medium is, for example, a hard disk drive, RAM, ROM, or semiconductor memory. The recording medium may be volatile or non-volatile. The compound basic data 651 indicates basic information about one or more compounds that have been tested. Each of the one or more compound detailed data 652 indicates detailed information about that compound.

[0470] FIG. 52 is a diagram showing an example of the compound basic data 651.

[0471] The compound basic data 651 indicates, for each of one or more compounds, an ID that is compound identification information for that compound, the composition formula of that compound, the process conditions for that compound, one or more properties of that compound, and the crystalline phase of that compound. For example, the compound basic data 651 shown in FIG. 52 indicates the ID "000001-00001-001" of an experimental compound and the composition formula of that compound "Li １．４５ La ０．０４５ Ti １．１ Al ０．００５ O ３", process conditions "firing temperature: 400°C, firing time: 3 hours", properties of the compound "Property 1: 2.349 eV, Property 2: 213°C", and the crystalline phase of the compound "cubic-perovskite". The composition of the compound, including its composition formula and process conditions, is defined by variables (M3, M3', M4, M4', a, b, Temp, Time, x, y) = (La, Al, Ti, Zr, 0.05, 0.1, 400, 3, 0.1, 0).

[0472] The ID, which is compound identification information, may be the name, code, or number string of the compound, as long as it is information that can identify the compound. The ID: 000001-00001-001 in Figure 52 is composed of three levels of numbers (000001, 00001, 001). For example, the IDs are registered so that the first level numbers "000001" to "000005" are assigned to the structures of five compounds with different composition ranges of the elements Zr and Ti contained in the compound.

[0473] The above-mentioned property 1 of the compound is, for example, the band gap, and the above-mentioned property 2 is the heat resistance temperature. Property 1, property 2, and the crystalline phase are each treated as experimental property values. Note that in the example shown in FIG. 52 , the band gap, the heat resistance temperature, and the crystalline phase are each treated as experimental property values, but this is not limited thereto, and any property depending on the compound may be treated as the experimental property value. For example, for a compound used as a battery material, electrical conductivity may be treated as the experimental property value, and for a compound used as a thermoelectric conversion material, a thermoelectric conversion figure of merit may be treated as the experimental property value.

[0474] FIG. 53 is a diagram showing an example of the compound detailed data 652.

[0475] The compound detail data 652 includes an ID, which is compound identification information of an experimented compound, and associated information of the compound. The associated information is so-called metadata, and includes, for example, the name of the researcher or experimenter who conducted the experiment on the compound, the registration date of the associated information, the temperature and humidity at the time the experiment was conducted, process information, and measurement data.

[0476] The process information indicates the raw materials used in the synthesis or production of the compound and the actual composition of the compound. １．４５１ La ０．０４５０ Ti １．１０２ Al ０．００５ O ３．０１ The process information may also include the experimental procedure for the compound, the shape and size of the crucible, the type of equipment used in the experiment, and the serial number of the equipment.

[0477] The measurement data is data obtained by a measuring instrument to derive experimental characteristic values, and indicates, for example, electrical resistance, differential scanning calorimetry (DSC), X-ray diffraction (XRD), etc. Note that the measurement data is not limited to these, and may indicate any data depending on the compound.

[0478] In this embodiment, the experimental database 650 stores the compound basic data 651 and the compound detailed data 652, but it is sufficient to store only the compound basic data 651. In this embodiment, storing the compound detailed data 652 in the experimental database 650 can more appropriately support material development. For example, if the compound detailed data 652 indicates the name of the researcher or experimenter, the registration date, the temperature, and the humidity, the credibility of the experimental data can be improved. Furthermore, if the compound detailed data 652 indicates process information, it is expected that more specific material properties can be searched for and the value of the experimental data can be increased. Furthermore, if the compound detailed data 652 indicates measurement data, the raw data of the experimental characteristic values ​​can be easily understood.

[0479] [Specific example of map] Figures 54 and 55 are diagrams showing examples of maps according to this embodiment. Note that Figure 56 shows a legend for the map in Figure 55. That is, [A] to [R] in Figure 55 are composition formulas associated with A to R shown in Figure 56.

[0480] 54 and 55 , an image map Mb is shown for each combination of choice data for each of the four categorical variables (M3, M3′, M4, M4′). This image map Mb is a map that shows, for each combination of choice data for each of the four variables (x, y, a, b), the property evaluation value of a compound having a configuration represented by that combination. The image map Mb is composed of an array (or matrix) of multiple image element maps Ma. Continuous variables x and y are assigned to the two coordinate axes of the image element map Ma, and discrete variables a and b are assigned to the two coordinate axes of the array of image element maps Ma (i.e., the image map Mb). Note that the variable assignment method is not limited to this. Property evaluation values ​​obtained for all combinations of choice data included in the search range may be displayed as is. Property evaluation values ​​for combinations other than all combinations included in the search range may be interpolated. Like predicted property values, these property evaluation values ​​are indicated by color or color shading.

[0481] Here, Figure 54 shows image maps Mb corresponding to each of two combinations of four categorical variables, and Figure 55 shows image maps Mb corresponding to each of 18 combinations of the four categorical variables. The number of image element maps Ma included in the map of the present disclosure is determined by the number of combinations of option data for each of the two discrete variables a and b and the number of combinations of option data for each of the four categorical variables (i.e., elements). In other words, the wider the search range indicated in the search range information d14, the greater the number of image element maps Ma. Therefore, it is important for users to devise a display that is visually easy to understand.

[0482] As an example of the display, the user can select the composition formula "Li ２－３ａ－４ｂ (M3 １－ｘ M3' ｘ ) ａ (M4 １－ｙ M4' ｙ ) １＋ｂ O ３54 and 55, a case where a characteristic evaluation value is searched for regarding the structure of a compound expressed by the process conditions "firing method Pa, firing time Pb, firing temperature Pc." will be described.

[0483] The image processing unit 634 acquires the first information d10 from the evaluation display database 640 via the display method acquisition unit 633, and generates a map according to the map format indicated in the map array information d13 included in the first information d10. Specifically, the image processing unit 634 first selects four variables a, b, x, and y from the continuous variables x and y, the discrete variables a and b, and the process variables Pa, Pb, and Pc. This is because it is necessary to assign four variables to four coordinate axes, consisting of the two coordinate axes of the image map Mb and the two coordinate axes of the image element map Ma included in the image map Mb.

[0484] Next, the image processing unit 634 assigns continuous variables x and y to the two coordinate axes of the image element map Ma. As in the first embodiment, each of the continuous variables x and y can take one of 11 option data values ​​ranging from a minimum value of 0.0 to a maximum valu...

Claims

1. Obtain the predicted property values of each of a plurality of compounds, Obtain first display method information indicating a display method of the predicted property values, Generate a map showing the predicted property values of each of the plurality of compounds according to the first display method information, Generate and output an image including the map, The map has coordinate axes respectively indicating at least two variables among a plurality of variables used to represent the composition of the compound, An information display method.

2. In the information display method, further, Obtain a plurality of variables used to represent the composition of the compound and, for each of the plurality of variables, a plurality of option data indicating values or elements that the variable can take, In the obtaining of the predicted property values, For each combination of option data obtained by selecting one option data from the plurality of option data for each of the plurality of variables, obtain the predicted property value of a compound having a configuration corresponding to the combination, The information display method according to Claim 1.

3. When there are inactive variables that are variables other than the at least two variables used for the coordinate axes of the map among the plurality of variables, The first display method information Indicates, as a display method of the predicted property values, to display the predicted property values using the inactive variables, The information display method according to Claim 1.

4. The first display method information When there are inactive variables that are variables other than the at least two variables used for the coordinate axes of the map among the plurality of variables, indicates, as a display method of the predicted property values, substituting a first value, a second value, or each numerical value within a predetermined numerical range into the inactive variables, (a) When a first value is substituted into the inactive variable, In the generation of the map, Generate the map showing the predicted property values of each of the plurality of compounds having a configuration represented by using the inactive variable indicating the first value specified by the user, (b) When a second value is substituted into the inactive variable, In the generation of the map, Determine the second value so that the predicted property values shown in the map satisfy a predetermined condition, Generate the map showing the predicted property values of each of the plurality of compounds having a configuration represented by using the inactive variable indicating the determined second value, (c) When each numerical value within the predetermined numerical range is substituted into the inactive variable, In the generation of the map, For each position on the map, calculate the average value of the predicted property values of a plurality of compounds having a configuration expressed using at least two variables each indicating a numerical value corresponding to the position. For each position on the map, generate the map showing the average value of the predicted property values calculated for the position. Each of the inactive variables of the plurality of compounds for which the average value of the predicted property values is calculated indicates different numerical values within the predetermined numerical range. The information display method according to claim 1.

5. In the information display method, further, Obtain the experimental property value of each of one or more compounds that have been experimented on. In the generation of the image, Superimpose the experimental property value of each of the one or more compounds that have been experimented on at the position on the map corresponding to the configuration of the compound, and generate the image including the map on which the experimental property value is superimposed. The information display method according to claim 1.

6. In the information display method, further, Obtain second display method information indicating the display method of the experimental property value. In the generation of the image, Superimpose the experimental property value on the map according to the second display method information. The information display method according to claim 5.

7. The predicted property value is shown on the map by a first display mode that is a color or color shade corresponding to the predicted property value. When the experimental property value is superimposed on the map as a mark having a second display mode that is a color or color shade corresponding to the experimental property value. The second display method information is Showing, as the display method of the experimental property value, to match the scale of the second display mode for the experimental property value with the scale of the first display mode for the predicted property value. The information display method according to claim 6.

8. The second display method information is When a plurality of the experimental property values are superimposed on the map as overlapping marks, showing, as the display method of the experimental property value, a rule for defining the order of overlap of the marks. In the generation of the image, Overlap the marks of each of the plurality of experimental property values on the map according to the rule. The information display method according to claim 6.

9. The rule is (a) The larger the experimental property value, the closer the mark of the experimental property value is arranged to the front side. (b) The closer the experimental property value is to a predetermined value, the closer the mark of the experimental property value is arranged to the front side, or (c) arranging the mark of the characteristic experimental value closer to the front side as the characteristic experimental value is closer to the characteristic predicted value shown at the position on the map where the characteristic experimental values are superimposed; The information display method according to claim 8.

10. The second display method information is when there is no position on the map corresponding to the configuration of the compound having the characteristic experimental value then (a) superimposing the characteristic experimental value at a position on the map corresponding to the configuration closest to the configuration of the compound having the characteristic experimental value on the map, or (b) not superimposing the characteristic experimental value on the map, is shown as the display method of the characteristic experimental value, In the generation of the image, processing regarding the superimposition of the characteristic experimental value on the map is performed according to the second display method information. The information display method according to claim 6.

11. The second display method information is showing, as the display method of the characteristic experimental value, superimposing, on the map, the characteristic experimental values satisfying a predetermined condition among the characteristic experimental values of the one or more compounds that have been experimented, in a manner that emphasizes them more than the characteristic experimental values not satisfying the predetermined condition, In the generation of the image, according to the second display method information, the characteristic experimental values satisfying the predetermined condition are superimposed on the map in a manner that emphasizes them. The information display method according to claim 6.

12. The predetermined condition is (a) the characteristic experimental value is a characteristic experimental value obtained before a predetermined period from the present, (b) the characteristic experimental value is one of a predetermined number of characteristic experimental values obtained most recently, (c) the characteristic experimental value is equal to or greater than a predetermined second threshold value, or (d) the difference between the characteristic experimental value and the characteristic predicted value obtained for a compound having the same configuration as the compound having the characteristic experimental value is equal to or greater than a predetermined third threshold value or less than the third threshold value, The information display method according to claim 11.

13. The map includes a plurality of image element maps arranged in a matrix along each of a first coordinate axis and a second coordinate axis, Each of the plurality of image element maps has a third coordinate axis and a fourth coordinate axis, In the generation of the map, the first coordinate axis, the second coordinate axis, the third coordinate axis, and the fourth coordinate axis are respectively associated with a first variable, a second variable, a third variable, and a fourth variable among the plurality of variables, For each of the plurality of compounds, Among the plurality of image element maps, identify the image element map associated with the values of the first variable and the second variable used to represent the structure of the compound. Map the predicted property value of the compound to the position corresponding to the values of the third variable and the fourth variable used to represent the structure of the compound on the identified image element map. The information display method according to any one of claims 6 to 12.

14. In the generation of the image, For each of the one or more compounds that have been experimented with, Among the plurality of image element maps, if there is no image element map associated with the values of the first variable and the second variable used to represent the structure of the compound, instead of the image element map, identify the image element map associated with the values closest to the values of each of the first variable and the second variable, Superimpose the experimentally measured property value of the compound on the position corresponding to the values of the third variable and the fourth variable used to represent the structure of the compound on the identified image element map. The information display method according to claim 13.

15. In the information display method, further, Obtain position information indicating the position of the predicted property value or the experimentally measured property value on the map, In the generation of the image, Obtain compositional formula data regarding the compositional formula of the compound corresponding to the position indicated by the position information, and superimpose a compositional image indicating the compositional formula data on the map, The compositional formula data includes inactive variables associated with the compound having the predicted property value or the experimentally measured property value, The inactive variables are variables other than the at least two variables used for the coordinate axes of the map among the plurality of variables. The information display method according to claim 6.

16. The map includes a plurality of image element maps arranged in a matrix along each of a first coordinate axis and a second coordinate axis. Each of the plurality of image element maps has a third coordinate axis and a fourth coordinate axis. The first coordinate axis and the second coordinate axis are respectively associated with a first variable and a second variable among the plurality of variables. The third coordinate axis and the fourth coordinate axis are respectively associated with a third variable and a fourth variable among the plurality of variables. The information display method according to claim 1.

17. Obtain the predicted characteristic values of each of a plurality of compounds, Output an image including a map generated using the obtained predicted characteristic values, The map has coordinate axes respectively indicating at least two variables used to represent the composition of the compound, and is a map showing the predicted characteristic values of each of the plurality of compounds. Information display method.

18. A predicted value acquisition unit that obtains the predicted characteristic values of each of a plurality of compounds, A display method acquisition unit that obtains first display method information indicating the display method of the predicted characteristic values, An image processing unit that outputs an image including a map showing the predicted characteristic values of each of the plurality of compounds, generated according to the first display method information, The map has coordinate axes respectively indicating at least two of the plurality of variables used to represent the composition of the compound. Information display device.

19. Obtain the predicted characteristic values of each of a plurality of compounds, Obtain first display method information indicating the display method of the predicted characteristic values, Generate a map showing the predicted characteristic values of each of the plurality of compounds according to the first display method information, Cause a computer to generate and output an image including the map, The map has coordinate axes respectively indicating at least two of the plurality of variables used to represent the composition of the compound. Program.

20. Obtain the predicted characteristic values of each of a plurality of compounds, Cause a computer to execute outputting an image including a map generated using the obtained predicted characteristic values, The map has coordinate axes respectively indicating at least two variables used to represent the composition of the compound, and is a map showing the predicted characteristic values of each of the plurality of compounds. Program.