Design support system, learning data generation system, and trained model generation system

Abstract

To provide a design support system suitable for grasping a plurality of elements to be edited successively and the order of editing them. [Solution] A drawing creation support device 100 acquires the ID of a created or edited element, and estimates multiple pieces of element order information from the acquired element ID using a trained model trained based on training data including the created or edited element ID, multiple edited elements edited consecutively after the edited element and element order information related to the edit order, and evaluation values ​​related to the multiple edited elements, which are evaluation values ​​related to the quality or risk of the building. This makes it possible to identify multiple edited elements that are edited consecutively and their editing order. It also makes it possible to identify the element order according to the evaluation values ​​related to the quality or risk of the building.

Description

[Technical Field]

[0001] The present invention relates to a design support system, and more particularly to a design support system, a training data generation system, and a trained model generation system that are suitable for grasping multiple elements that are edited in succession and the order in which they are edited. [Background technology]

[0002] BACKGROUND ART Conventionally, as a technology for supporting design using AI (Artificial Intelligence), for example, the technology described in Patent Document 1 is known.

[0003] The technology described in Patent Document 1 assigns a reward R to the combination of a state S, which is determined depending on whether or not a design activity has been performed, and an action A, which is an activity that can be selected under the state, and builds a trained model by maximizing the value.The trained model is then used to infer the next action to be taken from the current state. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2022-56238 Summary of the Invention [Problem to be solved by the invention]

[0005] In architectural design, multiple related elements of a unit may be edited consecutively. For example, when designing a toilet, the toilet bowl and hand basin are designed consecutively. However, the technology described in Patent Document 1 infers the next action to be taken from the current state, which poses a problem in that it is not possible to grasp the multiple elements that are edited consecutively and the order in which they are edited.

[0006] Therefore, the present invention has been made with a focus on the unresolved issues of the conventional technology, and aims to provide a design support system, a training data generation system, and a trained model generation system that are suitable for grasping multiple elements that are edited in succession and the order in which they are edited. [Means for solving the problem]

[0007] [Invention 1] In order to achieve the above object, the design support system of Invention 1 comprises: an element information acquisition means for acquiring element information relating to an element that has been created or edited in design information; and an estimation means for estimating the element order information from the element information acquired by the element information acquisition means, using a trained model trained based on training data including element information relating to the element that has been created or edited, element order information relating to a plurality of elements that have been edited consecutively after the element in question and their editing order (hereinafter, "the plurality of elements and their editing order" will be referred to as "element order"), and evaluation values ​​relating to the plurality of elements, wherein the evaluation value is an evaluation value relating to the quality or risk of the design object.

[0008] With this configuration, the element information acquisition unit acquires element information, and the estimation unit estimates element order information from the acquired element information using the trained model.

[0009] Here, the trained model may be one that has been trained based on training data that includes at least element information, element order information, and evaluation values, and also includes one that has been trained based on training data that includes element information, element order information, evaluation values, and other information.

[0010] Furthermore, the element information acquisition means may, for example, input element information from an input device or the like, acquire or receive element information from an external terminal or the like, read element information from a storage device or storage medium or the like, or generate or calculate element information by information processing or the like. Therefore, acquisition includes at least input, acquisition, reception, reading (including search), generation, and calculation. The same concept of acquisition applies hereinafter.

[0011] Furthermore, the element information may be configured, for example, as the element itself, or as information for identifying the element (for example, link information such as a name, number, ID, code, or URL), or as feature information relating to an outline, statistics, or other features of the element. The element information may be configured, for example, as characters, numbers, figures, codes, symbols, images, sounds, or other information. The element information may also be configured as keywords relating to the element (for example, one or more keywords indicating part of the name of the element). The same applies hereinafter to the trained model generation system of Invention 10.

[0012] Furthermore, the element order information may be configured, for example, as the element order itself, or as information for identifying the elements and the edit order (e.g., link information such as a name, number, ID, code, or URL), or as feature information relating to an overview, statistics, or other features of the elements and the edit order. The element order information may be configured, for example, as characters, numbers, figures, codes, symbols, images, sounds, or other information. The element order information may also be configured as keywords relating to the elements and the edit order (e.g., one or more keywords indicating part of the name of the elements and the edit order). The same applies to the training data generation system of Invention 8 and the trained model generation system of Invention 10.

[0013] Furthermore, this system may be realized as a single device, apparatus, terminal, or other device, or as a network system in which multiple devices, apparatus, terminals, or other devices are communicatively connected. In the latter case, each component may belong to any of the multiple devices as long as they are communicatively connected. The same applies below to the training data generation system of Invention 8 and the trained model generation system of Invention 10.

[0014] [Invention 2] Furthermore, the design support system of Invention 2 is the design support system of Invention 1, wherein the evaluation value regarding the quality of the design object is an evaluation value regarding the safety, functionality, convenience, aesthetics, environmental friendliness, economic efficiency, or legal compliance of the design object.

[0015] [Invention 3] Furthermore, the design support system of Invention 3 is the design support system of Invention 1, wherein the evaluation value relating to the risk of the design object is an evaluation value relating to the risk of a decline in the quality of the design object, the risk of the design object collapsing or being damaged, the risk of the design object deteriorating, the risk of an increase in costs relating to the design object, a risk to the design object due to a natural disaster, a risk of a decrease in the asset value of the design object, or a risk of the design object causing a social problem.

[0016] [Invention 4] Furthermore, the design support system of Invention 4 is the design support system of any one of Inventions 1 to 3, wherein the trained model is trained to maximize the evaluation value based on training data including the element information, the element order information, and the evaluation value.

[0017] [Invention 5] Furthermore, the design support system of Invention 5 is the design support system of Invention 4, wherein the trained model is a first trained model trained to maximize an evaluation value based on training data including the element information, the element order information, and an evaluation value based on a first index indicating an index of value of the evaluation value, and a second trained model trained to maximize an evaluation value based on training data including the element information, the element order information, and an evaluation value based on a second index different from the first index. The The system includes two trained models, and is equipped with an index information acquisition means that acquires index information regarding the first index or the second index, and a trained model selection means that selects either the first trained model or the second trained model based on the index information acquired by the index information acquisition means, and the estimation means estimates the element order information using the trained model selected by the trained model selection means.

[0018] With this configuration, the index information acquisition means acquires index information, the trained model selection means selects a trained model based on the acquired index information, and the estimation means estimates element order information using the selected trained model.

[0019] Here, the first trained model may be one that has been trained based on training data that includes at least element information, element order information, and an evaluation value based on the first index, and may also be one that has been trained based on training data that includes element information, element order information, an evaluation value based on the first index, and other information.

[0020] Furthermore, the second trained model may be one that has been trained based on training data that includes at least element information, element order information, and an evaluation value based on the second index, and may also be one that has been trained based on training data that includes element information, element order information, an evaluation value based on the second index, and other information.

[0021] [Invention 6] Furthermore, the design support system of Invention 6 is the design support system of Invention 1, in which the multiple elements are elements that require human judgment to set or change, and the setting or change affects other elements.

[0022] [Invention 7] Furthermore, in the design support system of Invention 7, in the design support system of Invention 1, the design information is design information for designing a building.

[0023] [Invention 8] Meanwhile, in order to achieve the above object, a training data generation system of Invention 8 includes: a calculation means for calculating an evaluation value of element order information based on first design information including element order information regarding a plurality of consecutively edited elements and their editing order (hereinafter, "a plurality of elements and their editing order" will be referred to as "element order") and having a first evaluation value set therefor, and second design information including element order information regarding the element order and having a second evaluation value set therefor; and a generation means for generating training data including the element order information included in the first design information and the second design information and the evaluation value calculated by the calculation means.

[0024] With this configuration, the generating means calculates an evaluation value based on the first design information and the second design information, and the generating means generates learning data including the element order information included in the first design information and the second design information and the calculated evaluation value.

[0025] [Invention 9] Furthermore, the training data generation system of Invention 9 is the training data generation system of Invention 8, wherein the calculation means calculates an evaluation value of the element order information included in the first design information based on the first evaluation value, calculates an evaluation value of the element order information included in the second design information based on the second evaluation value, and calculates an evaluation value of the element order information commonly included in the first design information and the second design information based on the first evaluation value and the second evaluation value.

[0026] With this configuration, the calculation means calculates an evaluation value of the element order information included in the first design information based on the first evaluation value, calculates an evaluation value of the element order information included in the second design information based on the second evaluation value, and calculates an evaluation value of the element order information commonly included in the first design information and the second design information based on the first evaluation value and the second evaluation value.

[0027] [Invention 10] Meanwhile, in order to achieve the above object, the trained model generation system of Invention 10 includes a generation means for generating a trained model by assigning an evaluation value to a combination of element information about an element that has been created or edited and element order information about multiple elements that have been edited consecutively after the element and their editing order (hereinafter, "multiple elements and their editing order" are referred to as "element order"), and performing training so as to maximize the evaluation value, wherein the evaluation value is an evaluation value related to the quality or risk of the design object.

[0028] With this configuration, an evaluation value is assigned to the combination of element information and element order information, and learning is performed so as to maximize the evaluation value, thereby generating a trained model. [Effects of the Invention]

[0029] As described above, the design support system of Invention 1 makes it possible to grasp multiple elements that are edited consecutively and the order in which they are edited. It also makes it possible to grasp the order of elements according to the evaluation values ​​related to the quality or risk of the object of design.

[0030] Furthermore, according to the design support system of invention 4, it is possible to grasp the element order with high evaluation values ​​regarding the quality or risk of the design object.

[0031] Furthermore, according to the design support system of invention 5, it is possible to grasp the order of elements with high evaluation values ​​based on the first index or the second index.

[0032] Furthermore, according to the design support system of Invention 6, it is possible to grasp the element order taking into consideration the relationship between elements whose setting or change requires human judgment and other elements that are affected by the setting or change.

[0033] On the other hand, according to the learning data generation system of invention 8, learning data including element order information and its evaluation values ​​can be generated based on design information in which evaluation values ​​are set.

[0034] On the other hand, according to the trained model generation system of Invention 10, a trained model can be obtained that provides an element order with a high evaluation value. [Brief explanation of the drawings]

[0035] [Figure 1] 1 is a diagram illustrating a hardware configuration of a drawing creation support device 100. FIG. [Figure 2] FIG. 10 is a diagram showing the structure of CAD data for a plan view. [Figure 3] FIG. 10 is a diagram showing the structure of edit history data. [Figure 4] 10 is a flowchart showing a learning data generation process. [Figure 5] FIG. 2 is a diagram illustrating a structure of learning data. [Figure 6] 10 is a flowchart illustrating a trained model generation process. [Figure 7] FIG. 1 is a block diagram showing the process of generating and using a trained model. [Figure 8] 10 is a flowchart showing an element order information estimation process. [Figure 9] This is a plan showing the expected delivery of a piano. [Figure 10] FIG. 10 is a diagram showing the structure of edit history data. [Figure 11] FIG. 2 is a diagram illustrating a structure of learning data. [Figure 12] 10 is a flowchart showing an element order information estimation process. [Figure 13] 10 is a diagram for explaining a case where an evaluation value of element order information is calculated from an evaluation value of a building. FIG. [Figure 14] This is a diagram showing three trained models 460 to 464 connected in parallel. [Figure 15] This is a diagram showing three trained models 470 to 474 connected in series. [Figure 16] 1 is a block diagram showing a configuration of a network system according to an embodiment of the present invention; [Figure 17] FIG. 1 is a functional block diagram of a generation AI server 120. [Figure 18] 10 is a flowchart showing a learning data registration process. [Figure 19] 10 is a flowchart showing an element order information acquisition process. [Figure 20] 10 is a flowchart showing an element order information acquisition process. DETAILED DESCRIPTION OF THE INVENTION

[0036] [First embodiment] A first embodiment of the present invention will be described below, with Figs. 1 to 9 showing the present embodiment.

[0037] [Configuration of this embodiment] First, the configuration of this embodiment will be described. FIG. 1 is a diagram showing the hardware configuration of a drawing creation support device 100. As shown in FIG.

[0038] As shown in FIG. 1, the drawing creation support device 100 is composed of a CPU (Central Processing Unit) 30 that controls calculations and the entire system based on a control program, a ROM (Read Only Memory) 32 that stores the control program and the like for the CPU 30 in advance in a predetermined area, a RAM (Random Access Memory) 34 that stores data read from the ROM 32 and the calculation results required in the calculation process of the CPU 30, and an I / F (Interface) 38 that mediates the input and output of data to and from external devices, and these are connected to each other and capable of sending and receiving data by a bus 39, which is a signal line for transferring data.

[0039] External devices connected to the I / F 38 include an input device 40 consisting of a keyboard, mouse, etc. that can input data as a human interface, a storage device 42 that stores data, tables, etc. as files, and a display device 44 that displays a screen based on an image signal.

[0040] CAD (Computer Aided Design) software and BIM (Building Information Modeling) software (hereinafter collectively referred to as "CAD software") are installed in the storage device 42. CAD software is software that assists designers in creating drawings in response to their operations. When a request is made to start the CAD software, the CPU 30 starts a program for the CAD software stored in a predetermined area of ​​the ROM 32 and executes processing in accordance with the program. The designer can start the CAD software to create plan drawings, detailed floor plans, and other architectural drawings.

[0041] Next, the data structure of the storage device 42 will be described. The storage device 42 stores CAD data of architectural drawings such as plan drawings, detailed floor plans, and the like.

[0042] FIG. 2 is a diagram showing the structure of CAD data for a plan view. As shown in Figure 2, CAD data for a plan is data that constitutes a detailed drawing of the cross section of a building, and is configured as data that includes one or more elements that can be created or edited (hereinafter referred to as "edit elements"). CAD data for a plan is created by a designer using CAD software. A designer creates a plan by creating, setting, changing, or deleting (hereinafter referred to as "editing") edit elements in the CAD software. In the example of Figure 2, the edit elements for the floor, wall, and ceiling of the area labeled "internal corridor" are arranged, and the edit elements for the floor, wall, and ceiling of the area labeled "windbreak room" are arranged.

[0043] The same is true for CAD data of detailed floor plans and other architectural drawings, which are structured as data including one or more editing elements.

[0044] The storage device 42 stores, for each CAD data, edit history data that indicates the history of editing of edit elements. The CAD data reflects the final edit results related to the edit history data.

[0045] FIG. 3 is a diagram showing the structure of the edit history data. 3, the editing history data includes, for each edited element, element information 400 about the edited element and editing time 402 required to edit the edited element, in the order of editing. The element information 400 includes an element ID for identifying the edited element, the area that was the target of editing of the edited element, and the edited element itself. The editing time 402 can be calculated, for example, by subtracting the editing start time from the editing end time.

[0046] In the example in Figure 3, the first related group shows that the toilet edit elements "toilet bowl," "hand wash counter," "mirror," and "towel rack" were edited consecutively in that order. These edit elements were assigned element IDs "52," "53," "54," and "55," and the edit times were 35, 38, 70, and 94 minutes, respectively.

[0047] The second related group shows that the edit elements of the toilet, "exhaust fan," "lighting," "storage," and "outlet," were edited consecutively in that order. These edit elements were assigned element IDs "56," "57," "58," and "59," and the edit times were 94, 48, 42, and 74 minutes, respectively.

[0048] The third related group shows that the closet edit elements "shelf," "clothes rack," "drawer," and "basket" were edited consecutively in that order. These edit elements were assigned element IDs "60," "61," "62," and "63," and the editing times were 96, 74, 84, and 22 minutes, respectively.

[0049] The fourth related group shows that the kitchen editing elements "flooring," "walling," "counter," and "sink" were edited consecutively in that order. These editing elements were assigned element IDs "64," "65," "66," and "67," and took 42, 44, 35, and 32 minutes to edit, respectively.

[0050] The editing history data is used to create learning data, and therefore the storage device 42 stores a large number of pieces of editing history data that have been created in the past.

[0051] [Operation of this embodiment] Next, the operation of this embodiment will be described. [Learning data generation process] FIG. 4 is a flowchart showing the learning data generation process.

[0052] The learning data generation process is a process for generating learning data, and when executed by the CPU 30, the process proceeds to step S100 as shown in FIG.

[0053] In step S100, unprocessed editing history data is obtained from the storage device 42, and the process proceeds to step S102, where the variable n is set to "2", and the process proceeds to step S104.

[0054] In steps S104 to S108, the element ID of the edited element that has been created or edited (hereinafter abbreviated as "created or edited element ID"), the n edited elements (the number indicated by the value of variable n) that were edited consecutively after that edited element, their edit order, and their edit time are obtained from the edit history data obtained in step S100. Hereinafter, multiple edited elements and their edit order may be referred to as the "element order." Using Figure 3 as an example, a case where the value of variable n is "2" will be described. In the edit history data in Figure 3, the rows are arranged in the order of editing.

[0055] For the second row, the element IDs of the previously edited elements, "01" to "51", are acquired as the created or edited element IDs. Because the value of variable n is "2", "toilet" and "hand wash counter" are acquired as the edited elements, and "35" and "38" are acquired as the edited times.

[0056] When the third row is targeted, the element IDs of the previous edited elements, "01" to "52", are acquired as the created or edited element IDs. Because the value of variable n is "2", "hand washing counter" and "mirror" are acquired as the edited elements, and "38" and "70" are acquired as the edited times.

[0057] Next, the process proceeds to step S110, where the evaluation value is calculated by multiplying the sum of the editing times obtained in step S108 by "-1." In the example of the second line above, the editing times "35" and "38" are obtained, so the evaluation value is calculated as (35 + 38) x -1 = -73. The reason for multiplying by "-1" is to set a higher evaluation value the shorter the editing time.

[0058] Next, the process proceeds to step S112, where the element ID and element order obtained in steps S104 to S108 and the evaluation value calculated in step S110 are associated and registered in the learning data, the process proceeds to step S114, where "1" is added to the value of variable n, and the process proceeds to step S116.

[0059] In step S116, it is determined whether the value of variable n is greater than "4", and if it is determined that the value is less than or equal to "4" (NO), the process proceeds to step S104. Then, the processes of steps S104 to S114 are repeatedly executed until the value of variable n becomes "4".

[0060] Using FIG. 3 as an example, the processing of steps S104 to S110 will be described for the case where the value of the variable n is "3."

[0061] When the second row is targeted, the element IDs of the previously edited elements, "01" to "51," are obtained as the created or edited element IDs. Because the value of variable n is "3," the edited elements are "toilet," "hand washing counter," and "mirror," and the edit times are "35," "38," and "70." The evaluation value is calculated as (35 + 38 + 70) x -1 = -143.

[0062] When the third row is targeted, the element IDs of the previous edited elements, "01" to "52," are acquired as the created or edited element IDs. Since the value of variable n is "3," the edited elements acquired are "hand washing counter," "mirror," and "towel rack," and the edit times are "38," "70," and "94." The evaluation value is calculated as (38 + 70 + 94) x -1 = -202.

[0063] Also, using FIG. 3 as an example, the processing of steps S104 to S110 will be described for the case where the value of the variable n is "4."

[0064] When the second row is targeted, the element IDs of the previously edited elements, "01" to "51," are acquired as created or edited element IDs. Because the value of variable n is "4," the edited elements acquired are "toilet," "hand wash counter," "mirror," and "towel rack," and the edit times acquired are "35," "38," "70," and "94." The evaluation value is calculated as (35 + 38 + 70 + 94) x -1 = -237.

[0065] When the third row is targeted, the element IDs of the previous edited elements, "01" to "52," are acquired as the created or edited element IDs. Because the value of variable n is "4," the edited elements acquired are "hand washing counter," "mirror," "towel rack," and "exhaust fan," and the edit times acquired are "38," "70," "94," and "94." The evaluation value is calculated as (38 + 70 + 94 + 94) x -1 = -296.

[0066] On the other hand, if it is determined in step S116 that the value of variable n is greater than "4" (YES), the process proceeds to step S118 to determine whether processing of steps S100 to S116 has been completed for all editing history data, and if it is determined that processing has been completed for all editing history data (YES), the process proceeds to step S120.

[0067] In step S120, the learning data in which the element IDs and other information have been registered in step S112 is stored in the storage device .

[0068] FIG. 5 is a diagram illustrating the structure of the learning data. 5, each row of the learning data includes a created or edited element ID 410, element order information 412, and an evaluation value 414. The element order information 412 includes the area to be edited for the edit element and the element order.

[0069] The second line of Fig. 5 shows the two edit elements edited consecutively after the edit element on the first line of Fig. 3, their edit order, and their evaluation values. The third line of Fig. 5 shows the two edit elements edited consecutively after the edit element on the second line of Fig. 3, their edit order, and their evaluation values. The fourth line of Fig. 5 shows the two edit elements edited consecutively after the edit element on the third line of Fig. 3, their edit order, and their evaluation values. The sixth line of Fig. 5 shows the two edit elements edited consecutively after the edit element on the fifth line of Fig. 3, their edit order, and their evaluation values. The seventh line of Fig. 5 shows the two edit elements edited consecutively after the edit element on the sixth line of Fig. 3, their edit order, and their evaluation values. The eighth line of Fig. 5 shows the two edit elements edited consecutively after the edit element on the seventh line of Fig. 3, their edit order, and their evaluation values.

[0070] 5 shows the three edited elements edited in succession after the edited element on the first line of Fig. 3, their edit order, and evaluation value, while line 11 of Fig. 5 shows the three edited elements edited in succession after the edited element on the second line of Fig. 3, their edit order, and evaluation value, while line 13 of Fig. 5 shows the three edited elements edited in succession after the edited element on the fifth line of Fig. 3, their edit order, and evaluation value, while line 14 of Fig. 5 shows the three edited elements edited in succession after the edited element on the sixth line of Fig. 3, their edit order, and evaluation value, respectively.

[0071] In addition, line 16 of Figure 5 shows the four edited elements that were edited consecutively after the edited element on line 1 of Figure 3, their edit order, and evaluation value, and line 18 of Figure 5 shows the four edited elements that were edited consecutively after the edited element on line 5 of Figure 3, their edit order, and evaluation value.

[0072] [Trained model generation process] FIG. 6 is a flowchart showing the trained model generation process.

[0073] FIG. 7 is a block diagram showing the process of generating and using a trained model. The trained model generation process is a process executed to generate a trained model, and when executed by the CPU 30, the process proceeds to step S200 to execute a training data analysis process, as shown in Fig. 6. In the training data analysis process, training data is read from the storage device 42, and created or edited element IDs 410, element order information 412, and evaluation values ​​414 are extracted from the read training data.

[0074] Next, the process proceeds to step S202. In step S202, as shown in FIG. 7, a training dataset is generated based on the information extracted in step S200. The process proceeds to step S204, where the generated training dataset is input to a training program, and a trained model is generated by the training program. The training program includes pre-training parameters and hyperparameters, and performs training based on the input training dataset and hyperparameters to update the pre-training parameters. As a training method, for example, reinforcement learning (e.g., supervised reinforcement learning, imitation learning) can be adopted. In reinforcement learning, an evaluation value related to the editing of the multiple edit elements is assigned to a combination of an already created or edited element ID, multiple edit elements edited consecutively after the edit element, and element order information related to the editing order, and training is performed to maximize the evaluation value. The trained model is then output as the training result.

[0075] The trained model is trained based on created or edited element IDs 410, element order information 412, and evaluation value 414 so as to maximize evaluation value 414. The trained model includes trained parameters in which pre-trained parameters have been updated through training, and an inference program. The inference program inputs created or edited element IDs, estimates element order information from the input element IDs based on the trained parameters, and outputs the estimated element order information. Note that the relationship between the input element IDs and the output element order information is determined by AI training, and therefore, although it shows a similar trend to the content of past training data, there is ambiguity in that it does not necessarily match exactly. However, this ambiguity can be reduced by adjusting the amount of training data and the training accuracy.

[0076] Next, the process proceeds to step S206, where the trained model generated in step S204 is stored in the storage device 42, and the series of processes ends.

[0077] [Element order information estimation process] FIG. 8 is a flowchart showing the element order information estimation process.

[0078] The element order information estimation process is executed in response to a request from a designer or other user, and when executed by the CPU 30, the process first proceeds to step S300 as shown in FIG.

[0079] In step S300, the created or edited element ID is acquired from the CAD data currently being edited, and the process proceeds to step S302.

[0080] In step S302, using the trained model in the storage device 42, multiple pieces of element order information with different edit elements or different edit orders are estimated from the element IDs acquired in step S300. The estimation is performed by inputting the element IDs into the trained model and acquiring the element order information output from the trained model. For the trained model, multiple outputs may be acquired from one input, or multiple outputs may be acquired by repeating one input and one output multiple times.

[0081] Next, the process proceeds to step S304, where one of a plurality of element orders is displayed on the display device 44 according to one of a plurality of display rules based on the plurality of element order information estimated in step S302. The display rule to be used may be set, for example, by a designer or other user, or may be set according to a predetermined algorithm.

[0082] The first display rule is to display the element order with the largest number of edit elements among the multiple estimated element orders.

[0083] The second display rule is to display an element order that has a related grouping among the multiple estimated element orders. For example, if the multiple element orders estimated from the current editing state are (1) A, B, (2) A, B, C, (3) A, B, C, D, (4) A, B, C, D, E, and (5) A, B, C, D, E, F, and A to D are a related grouping among them, according to the second display rule, one of (3) to (5) is displayed. Whether or not a grouping is related can be determined by, for example, identifying two or more edited elements and their edit orders among A to F that appear more than a predetermined number of times in the learning data. The same applies to the third and fourth display rules below.

[0084] The third display rule is to display the same element order as a related group of the estimated element orders. For example, if the element orders estimated from the current editing state are (1) to (5) above, the third display rule displays (3).

[0085] The fourth display rule is to display related groups among the estimated element sequences and related groups from the element sequence including other edited elements. For example, if the element sequence estimated from the current edit state is (1), (2), and (5) above, according to the fourth display rule, A to D are displayed from (5).

[0086] Next, the process proceeds to step S306, and if the designer creates or edits an edit element based on the displayed element order, the created or edited element ID including the element ID of the edit element is obtained from the CAD data currently being edited, and the process proceeds to step S308.

[0087] In step S308, based on the element order displayed in steps S304 and S314 and the element IDs acquired in step S306, it is determined whether the edited elements created or edited by the designer differ from the edited elements or edit order displayed in steps S304 and S314, and if it is determined that the edited results differ from the estimated results (YES), the process proceeds to step S310.

[0088] In step S310, the display rules used for display in steps S304 and S314 are changed. For example, the display rules are changed to the most appropriate ones so that the edited results match the estimated results. The timing for changing the display rules does not have to be every time one creation or edit is made, but may be every time multiple creations or edits are made.

[0089] Next, the process proceeds to step S312, where, similar to the process of step S302, multiple pieces of element order information are estimated from the element IDs acquired in step S306 using the trained model in the storage device 42, and the process proceeds to step S314.

[0090] In step S314, similar to the process in step S304, one of the plurality of element orders is displayed on the display device 44 based on the plurality of element order information estimated in step S312, and the process proceeds to step S316.

[0091] In step S316, it is determined whether or not the editing by the designer has finished, and if it is determined that the editing has finished (YES), the series of processes is ended.

[0092] On the other hand, if it is determined in step S316 that the editing by the designer has not been completed (NO), the process proceeds to step S306.

[0093] On the other hand, if it is determined in step S308 that the edited result matches the estimated result (NO), the process proceeds to step S312.

[0094] [When considering bringing in a piano] Next, the operation when a piano is brought in will be described.

[0095] Figure 9 is a plan showing how a piano will be brought in. If a designer wants to install a piano with a width of 120 mm in the plan view of Figure 9 in CAD software, he or she needs to shorten the width of the toilet to prevent the piano from interfering with the toilet wall when the piano is delivered. However, changing the toilet's width also requires editing other editable elements of the toilet. Therefore, after changing the toilet's width, the designer requests estimation of the element order to be edited for the toilet. Steps S300 to S304 are then performed to obtain created or edited element IDs from the currently edited CAD data, and the element order is estimated and displayed based on the obtained element IDs. For example, if the element order information in line 16 of Figure 5 is estimated, the following is displayed: "Toilet bowl → Hand washing counter → Mirror → Towel rack." As shown in Figure 9, if these editable elements are already included in the CAD data, for example, they are highlighted (e.g., displayed with a specific color or pattern) and the editing order is displayed by connecting them with arrows or the like. If they are not included in the CAD data, for example, they are ghost-displayed (e.g., ghost images of the editable elements are displayed in wireframe) and the editing order is displayed by connecting them with arrows or the like.

[0096] Then, when the designer creates or edits the edit element "toilet" based on the estimated and displayed element order, the next element order is estimated and displayed through steps S306 to S314. If the designer's editing result differs from the estimated result, the display rules are changed through step S310.

[0097] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, created or edited element IDs are acquired, and multiple pieces of element order information are estimated from the acquired element IDs using a trained model.

[0098] This makes it possible to grasp a plurality of editing elements that are edited consecutively and the editing order thereof.

[0099] Furthermore, in this embodiment, a plurality of pieces of element order information with different edit elements or edit orders are estimated, and one of a plurality of element orders is displayed based on the estimated plurality of pieces of element order information.

[0100] This makes it possible to grasp a plurality of edit elements to be edited successively and a plurality of candidates for the order of editing.

[0101] Furthermore, in this embodiment, the element order with the largest number of edited elements among the plurality of estimated element orders is displayed.

[0102] This allows the designer to carry out editing while imagining large editing units. Furthermore, in this embodiment, element orders that appear a predetermined number of times or more in the training data are identified as related groups, and element orders that have related groups among the multiple estimated element orders are displayed.

[0103] This allows the designer to carry out editing while imagining related groups.

[0104] Furthermore, in this embodiment, an element order that appears a predetermined number of times or more in the training data is identified as a related group, and an element order that is the same as the related group among the multiple estimated element orders is displayed.

[0105] This allows the designer to carry out editing while imagining related groups.

[0106] Furthermore, in this embodiment, element sequences that appear a predetermined number of times or more in the training data are identified as related groups, and from among the multiple estimated element sequences, related groups and element sequences that include other edited elements are displayed, which are related groups.

[0107] This allows the designer to carry out editing while imagining related groups.

[0108] Furthermore, in this embodiment, created or edited element IDs including the element IDs of edited elements created or edited by the designer are obtained based on the displayed element order, and multiple pieces of element order information are estimated from the obtained element IDs using a trained model.

[0109] As a result, the element order is estimated and displayed for the results of creation or editing based on the displayed element order, so that it is possible to grasp the multiple edit elements that are edited consecutively and their edit order.

[0110] Furthermore, in this embodiment, the created or edited element IDs, including the element IDs of the edited elements created or edited by the designer, are obtained based on the displayed element order, and the display rules are changed based on the displayed element order and the obtained element IDs.

[0111] This allows the display rules to be changed depending on the display result of the element order and the result of subsequent creation or editing.

[0112] Furthermore, in this embodiment, the trained model is trained to maximize the evaluation value based on training data including the created or edited element ID, multiple edited elements edited consecutively after that edited element and element order information regarding the order of editing, and evaluation values ​​regarding the editing of the multiple edited elements.

[0113] This allows the order of elements with high evaluation values ​​to be understood. Furthermore, in this embodiment, an evaluation value related to the editing of multiple edited elements is assigned to a combination of an element ID that has been created or edited, multiple edited elements that have been edited consecutively after that edited element, and element order information related to the order of editing, and a trained model is generated by performing training so that the evaluation value is maximized.

[0114] This makes it possible to obtain a trained model that produces an element order with a high evaluation value. Second Embodiment Next, a second embodiment of the present invention will be described. Figures 10 to 12 show this embodiment. Figures 3, 5, 6 and 9 are also used.

[0115] This embodiment differs from the first embodiment in that estimation is performed using a first trained model and a second trained model that have been trained using evaluation values ​​based on different indices. Only the differences from the first embodiment will be described below, and descriptions of overlapping parts will be omitted.

[0116] [Configuration of this embodiment] First, the configuration of this embodiment will be described. FIG. 10 is a diagram showing the structure of the edit history data.

[0117] The storage device 42 stores the edit history data of FIG. 10 in addition to the edit history data of FIG. 3 for each CAD data.

[0118] As shown in FIG. 10, the edit history data includes, for each edited element, element information 420 relating to that edit element and the number of edited items 422 required to edit that edit element, in the order of editing.

[0119] In the example of Figure 10, the first related group shows that the edit elements of the toilet, "toilet bowl," "hand wash counter," "mirror," and "towel rack," are edited consecutively in that order. These edit elements are assigned element IDs "52," "53," "54," and "55," and the number of edit items is 8, 5, 6, and 9, respectively.

[0120] The second related group shows that the edit elements of the toilet, "exhaust fan," "lighting," "storage," and "outlet," are edited consecutively in that order. These edit elements are assigned element IDs "56," "57," "58," and "59," and the number of edit items is 3, 7, 8, and 6, respectively.

[0121] The third related group shows that the closet edit elements "shelf," "hanger rail," "drawer," and "basket" are edited consecutively in that order. These edit elements are assigned element IDs "60," "61," "62," and "63," and the number of edit items is 8, 7, 2, and 4, respectively.

[0122] The fourth related group shows that the kitchen edit elements "flooring," "wall material," "counter," and "sink" are edited consecutively in that order. These edit elements are assigned element IDs "64," "65," "66," and "67," and the number of edit items is 7, 4, 3, and 8, respectively.

[0123] [Operation of this embodiment] Next, the operation of this embodiment will be described. [Learning data generation process] In the learning data generation process, learning data is generated based on the edit history data of FIG. 10, similarly to the generation of learning data of FIG.

[0124] FIG. 11 is a diagram showing the structure of the learning data. As shown in FIG. 11, the learning data includes, for each line, an element ID 430 that has been created or edited, element order information 432, and an evaluation value 434.

[0125] The second line of Fig. 11 shows two edit elements edited consecutively after the edit element on the first line of Fig. 10, their edit order, and evaluation value. The third line of Fig. 11 shows two edit elements edited consecutively after the edit element on the second line of Fig. 10, their edit order, and evaluation value. The fourth line of Fig. 11 shows two edit elements edited consecutively after the edit element on the third line of Fig. 10, their edit order, and evaluation value. The sixth line of Fig. 11 shows two edit elements edited consecutively after the edit element on the fifth line of Fig. 10, their edit order, and evaluation value. The seventh line of Fig. 11 shows two edit elements edited consecutively after the edit element on the sixth line of Fig. 10, their edit order, and evaluation value. The eighth line of Fig. 11 shows two edit elements edited consecutively after the edit element on the seventh line of Fig. 10, their edit order, and evaluation value.

[0126] 11 shows the three edited elements edited in succession after the edited element on the first line of Fig. 10, their edit order, and evaluation value, while line 11 of Fig. 11 shows the three edited elements edited in succession after the edited element on the second line of Fig. 10, their edit order, and evaluation value, while line 13 of Fig. 11 shows the three edited elements edited in succession after the edited element on the fifth line of Fig. 10, their edit order, and evaluation value, while line 14 of Fig. 11 shows the three edited elements edited in succession after the edited element on the sixth line of Fig. 10, their edit order, and evaluation value, respectively.

[0127] In addition, the 16th line of Figure 11 shows the four edited elements that were edited consecutively after the edited element on the first line of Figure 10, their edit order, and their evaluation value, and the 18th line of Figure 11 shows the four edited elements that were edited consecutively after the edited element on the fifth line of Figure 10, their edit order, and their evaluation value.

[0128] [Trained model generation process] In the trained model generation process, steps S200 to S204 are performed to generate a first trained model by performing training based on the training data in Fig. 5. The first trained model is the same as the trained model in the first embodiment.

[0129] In the trained model generation process, steps S200 to S204 are performed, and a second trained model is generated by performing training based on the training data in Fig. 11, similar to the generation of the first trained model. The second trained model is trained based on the created or edited element IDs 430, element order information 432, and evaluation values ​​434 so as to maximize the evaluation value 434.

[0130] Then, the process proceeds to step S206, where the first trained model and the second trained model generated in step S204 are stored in the storage device 42, and the series of processes is terminated.

[0131] [Element order information estimation process] FIG. 12 is a flowchart showing the element order information estimation process.

[0132] The element order information estimation process is a process that is executed in response to a request from a designer or other user, and when executed by the CPU 30, as shown in FIG. 12, first the process proceeds to step S330.

[0133] In step S330, index information regarding the first index "editing time" or the second index "number of edited items", which indicate an index of the value of the evaluation value, is acquired, and the process proceeds to step S332, where if the index related to the acquired index information is the first index, the first trained model is selected, and if the index related to the acquired index information is the second index, the second trained model is selected.

[0134] Next, the process proceeds to step S334, where the created or edited element ID is acquired from the CAD data currently being edited, and the process proceeds to step S336.

[0135] In step S336, the trained model selected in step S332 from the first trained model and the second trained model in the storage device 42 (hereinafter referred to as the "selected trained model") is used to estimate multiple pieces of element order information from the element IDs acquired in step S334. The estimation method is the same as the processing in step S302 in the first embodiment described above.

[0136] Next, the process proceeds to step S338, where one of the plurality of element orders is displayed on the display device 44 based on the plurality of element order information estimated in step S336, similar to the process of step S304, and the process proceeds to step S340.

[0137] In step S340, if the designer creates or edits an edit element based on the displayed element order, the created or edited element ID including the element ID of the edit element is obtained from the CAD data currently being edited, and the process proceeds to step S342.

[0138] In step S342, based on the element order displayed in steps S338 and S348 and the element IDs acquired in step S340, it is determined whether the edited elements created or edited by the designer differ from the edited elements or edit order displayed in steps S338 and S348, and if it is determined that the edited results differ from the estimated results (YES), the process proceeds to step S344.

[0139] In step S344, similar to the process in step S310, the display rules used for display in steps S338 and S348 are changed, and the process proceeds to step S346.

[0140] In step S346, similar to the processing in step S336, the selected trained model is used to estimate multiple pieces of element order information from the element IDs acquired in step S340, and the process proceeds to step S348.

[0141] In step S348, similar to the process in step S338, one of the plurality of element orders is displayed on display device 44 based on the plurality of element order information estimated in step S346, and the process proceeds to step S350.

[0142] In step S350, it is determined whether or not the editing by the designer has finished, and if it is determined that the editing has finished (YES), the series of processes is ended.

[0143] On the other hand, if it is determined in step S350 that the editing by the designer has not been completed (NO), the process proceeds to step S340.

[0144] On the other hand, if it is determined in step S342 that the edited result matches the estimated result (NO), the process proceeds to step S346.

[0145] [When considering bringing in a piano] Next, the operation when a piano is brought in will be described.

[0146] If a designer wants to install a piano with a width of 120 mm in the plan view of FIG. 9 in CAD software, the designer needs to shorten the width of the toilet to prevent the piano from interfering with the toilet wall when bringing the piano in. However, changing the width of the toilet will require editing of other elements of the toilet as well. Therefore, after changing the width of the toilet, the designer requests an estimation of the element order to be edited for the toilet. At this time, if the designer wants to obtain an element order that shortens the editing time, the designer selects "editing time" as the indicator, and the first trained model is selected through steps S330 to S332. Then, through steps S334 to S338, created or edited element IDs are obtained from the CAD data currently being edited, and the first trained model estimates and displays the element order from the obtained element IDs.

[0147] On the other hand, if an element order that reduces the number of edit items is desired, the designer selects "number of edit items" as the index, and the second trained model is selected through steps S330 to S332. Then, through steps S334 to S338, created or edited element IDs are acquired from the CAD data currently being edited, and the second trained model estimates and displays the element order from the acquired element IDs.

[0148] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, index information regarding the first index or the second index is acquired, either the first trained model or the second trained model is selected based on the acquired index information, created or edited element IDs are acquired, and multiple element order information is estimated from the acquired element IDs using the selected trained model.

[0149] This makes it possible to grasp the order of elements with high evaluation values ​​based on the first index or the second index.

[0150] [Third embodiment] Next, a third embodiment of the present invention will be described with reference to Figure 13.

[0151] This embodiment differs from the first and second embodiments in that it employs an evaluation value other than the editing time or the number of edited items. Only the differences from the first and second embodiments will be described below, and explanations of overlapping parts will be omitted.

[0152] [Configuration of this embodiment] First, the configuration of this embodiment will be described. The evaluation values ​​to be included in the learning data are evaluation values ​​relating to the quality or risk of buildings.

[0153] The quality of a building can be determined, for example, from (1) the safety of the building (e.g., structural safety, fire safety, and safety in use), (2) the functionality or convenience of the building (e.g., livability and comfort), (3) the aesthetics of the building (e.g., design), (4) the environmental friendliness of the building (e.g., reduced environmental impact and energy conservation), (5) the economic efficiency of the building (e.g., construction costs, maintenance costs, life cycle costs, and other costs), and (6) the legal compliance of the building (e.g., the degree to which it complies with laws and standards, or whether it complies at all).

[0154] Risks to buildings include, for example, (1) the risk of a decline in the quality of the building (for example, the above-mentioned safety, functionality, convenience, aesthetics, environmental friendliness, economic efficiency, or legal compliance), (2) the risk of the building collapsing or being damaged due to a lack of strength, durability, earthquake resistance, etc., (3) the risk of the building deteriorating due to aging, etc., (4) the risk of an increase in costs related to the building, (5) risks to the building due to fire, flood, wind damage, earthquake, or other natural disasters, (6) the risk of a decrease in the asset value of the building, and (7) the risk of the building causing social problems.

[0155] The quality or risk of a building can be evaluated as a numerical value within a predetermined range (for example, 0 to 10), which can be used as the evaluation value. It is also preferable to standardize the evaluation value for each index (average value 0, standard deviation 1) so that evaluation values ​​based on different indexes can be compared to the same extent. The evaluation value for the quality or risk of a building may be set for the element order information as in the first and second embodiments above, or may be set for the building to be designed. When setting an evaluation value for a building, the learning data requires an evaluation value corresponding to each element order information, so the evaluation value for the element order information related to the design of that building is calculated from the evaluation value for the building.

[0156] FIG. 13 is a diagram for explaining a case where an evaluation value of element order information is calculated from an evaluation value of a building.

[0157] 13, there are three CAD data 450 to 454, and one evaluation value is set for each CAD data as a whole. Note that the evaluation value is assumed to be standardized.

[0158] CAD data 450 is CAD data related to the design of building A, and an evaluation value of index 1 (e.g., safety of the building) is set for building A. For example, an evaluation value of "0.1" based on index 1 is set for CAD data 450.

[0159] CAD data 452 is CAD data related to the design of building B, and an evaluation value of index 2 (for example, functionality of the building) is set for building B. For example, for CAD data 452, an evaluation value based on index 2 of "0.2" is set.

[0160] CAD data 454 is CAD data related to the design of building C, and an evaluation value based on index 3 (e.g., building risk) is set for building C. For example, for CAD data 454, an evaluation value based on index 3 of "0.3" is set.

[0161] CAD data 450 to 454 contain multiple pieces of element order information. In step S110, an evaluation value of "0.1" is calculated for element order information contained only in CAD data 450, an evaluation value of "0.2" for element order information contained only in CAD data 452, and an evaluation value of "0.3" for element order information contained only in CAD data 454. Furthermore, an evaluation value of 0.1 + 0.2 = "0.3" is calculated for element order information contained in common by CAD data 450 and 452, an evaluation value of 0.2 + 0.3 = "0.5" for element order information contained in common by CAD data 452 and 454, and an evaluation value of 0.1 + 0.3 = "0.4" for element order information contained in common by CAD data 450 to 454. Furthermore, an evaluation value of 0.1 + 0.2 + 0.3 = "0.6" is calculated for element order information contained in common by CAD data 450 to 454.

[0162] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, an element ID that has been created or edited is obtained, and multiple pieces of element order information are estimated from the obtained element ID using a trained model trained based on training data including the created or edited element ID, multiple edited elements edited consecutively after the edited element and element order information related to the order of editing, and evaluation values ​​related to the multiple edited elements, which are evaluation values ​​related to the quality or risk of the building.

[0163] This allows the user to grasp the multiple edit elements that are edited consecutively and the order in which they are edited, as well as the order in which elements are edited according to the evaluation values ​​related to the quality or risk of the building.

[0164] Furthermore, in this embodiment, the trained model is trained to maximize the evaluation value based on training data including element IDs, element order information, and evaluation values.

[0165] This makes it possible to understand the order of elements with high evaluation values ​​regarding the quality or risk of a building.

[0166] In this embodiment, step S110 corresponds to the calculation means of invention 8 or 9, step S120 corresponds to the generation means of invention 8 or 10, steps S300, S306, S334, and S340 correspond to the element information acquisition means of invention 1, and steps S302, S312, S336, and S346 correspond to the estimation means of invention 1 or 5. Also, step S330 corresponds to the index information acquisition means of invention 5, step S332 corresponds to the model selection means of invention 5, CAD data corresponds to the design information of inventions 7 to 9, and created or edited element IDs correspond to the element information of inventions 1, 4, 5, or 10.

[0167] [Fourth embodiment] Next, a fourth embodiment of the present invention will be described. Fig. 14 shows this embodiment. Fig. 5 and Fig. 11 are also used.

[0168] This embodiment differs from the third embodiment in that element order information is estimated using multiple trained models that are each trained based on evaluation values ​​that are based on different indices. Only the differences from the first to third embodiments will be described below, and descriptions of overlapping parts will be omitted.

[0169] [Configuration of this embodiment] First, the configuration of this embodiment will be described. FIG. 14 is a diagram in which three trained models 460 to 464 are connected in parallel.

[0170] As shown in FIG. 14, the storage device 42 stores three trained models 460 to 464.

[0171] 5 or 11, learning data 1 is prepared in which an evaluation value based on a first index (for example, safety of a building) is set. The trained model 460 is trained based on the learning data 1. The trained model 460 receives an element ID as input, estimates one or more pieces of element order information having a high evaluation value based on the first index from the input element ID, and outputs the estimated element order information.

[0172] 5 or 11, learning data 2 is prepared in which an evaluation value based on a second index (for example, functionality of a building) is set. The trained model 462 is trained based on the learning data 2. The trained model 462 receives element IDs as input, estimates one or more pieces of element order information having a high evaluation value based on the second index from the input element IDs, and outputs the estimated element order information.

[0173] 5 or 11, learning data 3 is prepared in which an evaluation value based on a third index (e.g., building risk) is set. The trained model 464 is trained based on the learning data 3. The trained model 464 receives element IDs as input, estimates one or more pieces of element order information having a high evaluation value based on the third index from the input element IDs, and outputs the estimated element order information.

[0174] The method for generating the trained models 460 to 464 is the same as in the first and second embodiments.

[0175] In step S302, estimation is performed by inputting the element IDs acquired in step S300 into each of the trained models 460 to 464 and acquiring element order information output by the determination unit 466. The same applies to steps S312, S336, and S346.

[0176] The determination unit 466 determines one or more pieces of element order information from the element order information output by the trained models 460 to 464. The determination unit 466 may have the following configuration, for example.

[0177] In the first configuration, element order information is determined according to the number of times the element order information appears. For example, if element order information A and B are output from trained model 460, element order information A, B, and C are output from trained model 462, and element order information A is output from trained model 464, the number of times the element order information A to C appear will be "3," "2," and "1," respectively. Therefore, if it appears in the top one position, element order information A is output, and if it appears in the top two positions, element order information A and B are output.

[0178] In the second configuration, element order information is determined according to the number of occurrences of edited elements included in the element order information. For example, in the above example in the first configuration, if element order information A includes edited elements a and b, element order information B includes edited elements b and c, and element order information C includes edited elements a and c, the number of occurrences of edited elements a, b, and c will be "4," "5," and "3," respectively. If the number of occurrences of element order information A to C is calculated by adding up the number of occurrences of the edited elements included in that element order information, the number of occurrences of element order information A to C will be "9," "8," and "3," respectively. Therefore, if it is the top one, element order information A is output, and if it is the top two, element order information A and B are output.

[0179] Trained models 460 to 464 are configured to output evaluation values ​​and fitness values ​​along with element order information. A third configuration determines element order information according to evaluation values, etc. In the above example of the first configuration, if the evaluation values, etc. of element order information A to C are "0.2," "0.4," and "0.5," respectively, and the evaluation values, etc. of element order information A to C are corrected by multiplying them by the number of occurrences, the evaluation values, etc. of element order information A to C become "0.2 x 3 = 0.6," "0.4 x 2 = 0.8," and "0.5 x 1 = 0.5," respectively. Therefore, if the element order information is in the top one position, element order information B is output, and if the element order information is in the top two positions, element order information A and B are output.

[0180] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, multiple trained models, each trained based on evaluation values ​​based on different indices, are connected in parallel, element IDs are input to each trained model, and element order information is determined from the output results of each trained model, thereby estimating element order information.

[0181] This makes it possible to grasp the element order according to the evaluation values ​​based on a plurality of different indices.

[0182] Fifth Embodiment Next, a fifth embodiment of the present invention will be described. Fig. 15 shows this embodiment. Fig. 5 and Fig. 11 are also used.

[0183] This embodiment differs from the third embodiment in that element order information is estimated using multiple trained models that are each trained based on evaluation values ​​that are based on different indices. Only the differences from the first to third embodiments will be described below, and descriptions of overlapping parts will be omitted.

[0184] [Configuration of this embodiment] First, the configuration of this embodiment will be described. FIG. 15 is a diagram showing three trained models 470 to 474 connected in series.

[0185] As shown in FIG. 15, the storage device 42 stores three trained models 470 to 474.

[0186] 5 or 11, learning data 1 is prepared in which an evaluation value based on a first index (for example, safety of a building) is set. The trained model 470 is trained based on the learning data 1. The trained model 470 receives an element ID as input, estimates one or more pieces of element order information having a high evaluation value based on the first index from the input element ID, and outputs the estimated element order information.

[0187] 5 or 11, element IDs 410 and 430 are deleted, and training data 2 is prepared by setting an evaluation value based on a second index (for example, functionality of a building). Trained model 472 is trained based on training data 2. Trained model 472 receives the element order information output by trained model 470, estimates one or more pieces of element order information with a high evaluation value based on the second index from the received element order information, and outputs the estimated element order information.

[0188] 5 or 11, element IDs 410 and 430 are deleted, and training data 3 is prepared by setting an evaluation value based on a third index (e.g., building risk). Trained model 474 is trained based on training data 3. Trained model 474 receives the element order information output by trained model 472, estimates one or more pieces of element order information with a high evaluation value based on the third index from the received element order information, and outputs the estimated element order information.

[0189] The method for generating the trained models 470 to 474 is the same as in the first and second embodiments.

[0190] In step S302, estimation is performed by inputting the element IDs acquired in step S300 into the trained model 470 and acquiring element order information output by the trained model 474. The same applies to steps S312, S336, and S346.

[0191] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, multiple trained models, each trained based on evaluation values ​​based on different indices, are connected in series, element IDs are input to the trained model at the front stage, and element order information is estimated by obtaining the element order information output by the trained model at the final stage.

[0192] This makes it possible to grasp the element order according to the evaluation values ​​based on a plurality of different indices.

[0193] Sixth Embodiment Next, a sixth embodiment of the present invention will be described. Figures 16 to 19 show this embodiment. Figures 5 and 9 are also used.

[0194] This embodiment differs from the first embodiment in that a large language model is used. Only the differences from the first embodiment will be described below, and descriptions of overlapping parts will be omitted.

[0195] [Configuration of this embodiment] First, the configuration of this embodiment will be described. FIG. 16 is a block diagram showing the configuration of a network system according to this embodiment.

[0196] As shown in FIG. 16, the drawing creation support device 100 and a generation AI server 120 that generates answer information in response to a request using an AI (Artificial Intelligence) model are connected to the Internet 199 so as to be able to communicate with each other.

[0197] [Generation AI Server 120] Next, the configuration of the generation AI server 120 will be described. Like the drawing creation support device 100, the generation AI server 120 has a hardware configuration similar to that of a general computer in which a CPU, ROM, RAM, I / F, etc. are connected by a bus, and is configured as, for example, a cloud server.

[0198] FIG. 17 is a functional block diagram of the generation AI server 120. As shown in Figure 17, the generation AI server 120 is configured to have multiple AI models 50, an AI model control unit 52 that controls the AI ​​models 50, and a knowledge base 54 that registers data that the AI ​​models 50 refer to for inference.

[0199] The AI ​​model 50 is an AI model trained on a large data set and is a highly versatile model capable of performing a variety of tasks. For example, a large-scale language model can be used as the AI ​​model 50. A large-scale language model is a deep learning model that pre-trains a language model, which models human-spoken language based on its occurrence probability, from a massive amount of data. When a prompt is input, the large-scale language model statistically infers the probability of generating the next word from the sentence included in the input prompt and outputs the inference result. For example, publicly known technologies described on the internet sites "https: / / chatgpt-lab.com / n / n418d3aa56f0b" and "https: / / agirobots.com / chatgpt-mechanism-and-problem / " can be used as the large-scale language model. More specifically, for example, Titan Text G1 - Express, Titan Text G1 - Lite, Titan Image Generator G1, Titan Embeddings G1 - Text, Titan Embeddings Text V2, Titan Multimodal Embeddings G1, Claude, Claude Instant, Claude 3 Sonnet, Claude 3 Haiku, Claude 3 Opus, Jurassic-2 Mid, Jurassic-2 Ultra, Command, Command Light, Command R, Command R+, Embed English, Embed Multilingual, Llama 2 Chat 13B, Llama 2 Chat 70B, Llama 2 13B, Llama 2 70B, Llama 3 8b Instruct, Llama 3 70b Instruct, Mistral 7B Instruct, Mixtral 8X7B Instruct, Mistral Large, and Stable Diffusion XL can be adopted.

[0200] The AI ​​model control unit 52 selects one of the multiple AI models 50 to be used for inference in response to a selection request from the request processing unit 58. Furthermore, when a reference request is input from the request processing unit 58, the AI ​​model control unit 52 causes the selected AI model 50 (hereinafter referred to as the "selected AI model") to refer to the data in the knowledge base 54 in response to the input reference request. Furthermore, when a prompt is input from the request processing unit 58, the input prompt is input to the selected AI model. Then, when an execution request is input from the request processing unit 58, the AI ​​model control unit 52 causes the selected AI model to execute inference in response to the input execution request, obtains an inference result from the selected AI model, and outputs the obtained inference result to the request processing unit 58.

[0201] Learning data can be registered in the knowledge base 54. Information registered in the knowledge base 54 is in a data format (for example, vector data) that can be referenced by the AI ​​model 50.

[0202] The generation AI server 120 is further configured to include a request receiving unit 56 that receives requests, a request processing unit 58 that processes the requests received by the request receiving unit 56, and an answer information sending unit 60 that sends answer information to the request received by the request receiving unit 56 to the drawing creation support device 100.

[0203] The request receiving unit 56 receives a request for generating answer information from the drawing creation support device 100 and outputs the received request to the request processing unit 58. The request includes (1) an element ID that has already been created or edited, (2) a generation request to generate element order information regarding a plurality of edit elements to be edited successively after the edit element that has already been created or edited and the order in which they are edited, (3) a selection request to select an AI model 50, and (4) a reference request to reference learning data in the knowledge base 54. (3) and (4) are not essential and are included additionally.

[0204] When a request received by the request receiving unit 56 includes a selection request or a reference request, the request processing unit 58 outputs the selection request or the reference request to the AI ​​model control unit 52. Furthermore, based on the request received by the request receiving unit 56, the request processing unit 58 generates a prompt that instructs the AI ​​model 50. The prompt, for example, requests the AI ​​model 50 to generate, based on the created or edited element IDs, a plurality of edited elements to be edited consecutively after the created or edited edited element and element order information regarding the edit order, with an evaluation value equal to or greater than a predetermined value. The generated prompt and execution request are then output to the AI ​​model control unit 52. When an inference result is input from the AI ​​model control unit 52 in response to the execution request, the request processing unit 58 outputs the input inference result to the answer information sending unit 60.

[0205] The answer information sending unit 60 sends the answer information including the inference result input from the request processing unit 58 to the drawing creation support device 100.

[0206] The generation AI server 120 further includes a request receiving unit 62 that receives requests, and a learning data registration unit 64 that registers learning data in the knowledge base 54.

[0207] The request receiving unit 62 receives a request for registering learning data from the drawing production support device 100, and outputs the received request to the learning data registration unit 64. The request includes (1) learning data.

[0208] The learning data registration unit 64 stores the learning data included in the request received by the request receiving unit 62 in storage (not shown) and converts it into a data format (e.g., vector data) that can be referenced by the AI ​​model 50. Vector data can be generated by a technique (embedding) that converts data including characters, images, audio, etc. into a numerical vector. The converted learning data is then registered in the knowledge base 54. The AI ​​model control unit 52 causes the AI ​​model 50 to reference the learning data in response to a reference request from the request processing unit 58.

[0209] [Operation of this embodiment] Next, the operation of this embodiment will be described. [Learning data registration process] FIG. 18 is a flowchart showing the learning data registration process.

[0210] The learning data registration process is executed in response to a request from a designer or other user, and when executed by the CPU 30, the process first proceeds to step S400 as shown in FIG.

[0211] In step S400, the learning data of FIG. 5 is acquired from the storage device 42, and the process proceeds to step S402.

[0212] In step S402, a request for registering the learning data is sent to the generation AI server 120. The request includes (1) the learning data acquired in step S400.

[0213] When the process of step S402 ends, the series of processes ends. [Element order information acquisition process] FIG. 19 is a flowchart showing the element order information acquisition process.

[0214] The element order information acquisition process is executed in response to a request from a designer or other user, and when executed by the CPU 30, the process first proceeds to step S500 as shown in FIG.

[0215] In step S500, the created or edited element ID is acquired from the CAD data currently being edited, and the process proceeds to step S502.

[0216] In step S502, a request for generating answer information based on the element ID acquired in step S500 is sent to the generation AI server 120. The request includes (1) the element ID acquired in step S500, (2) a generation request to generate element order information regarding multiple edit elements to be edited consecutively after an edit element that has already been created or edited and the order in which they are edited, where the edit elements or edit orders are different, (3) a selection request to select a specific AI model 50, and (4) a reference request to reference the learning data in the knowledge base 54.

[0217] Next, proceed to step S504 to receive answer information from the generation AI server 120, proceed to step S506 to display one of the multiple element orders on the display device 44 based on the multiple element order information contained in the received answer information, similar to the processing of step S304, and proceed to step S508.

[0218] In step S508, if the designer creates or edits an edit element based on the displayed element order, the created or edited element ID including the element ID of the edit element is obtained from the CAD data currently being edited, and the process proceeds to step S510.

[0219] In step S510, based on the element order displayed in steps S506 and S518 and the element ID acquired in step S508, it is determined whether the edited elements created or edited by the designer differ from the edited elements or edit order displayed in steps S506 and S518, and if it is determined that the edited results differ from the inference results (YES), the process proceeds to step S512.

[0220] In step S512, similar to the process in step S310, the display rules used for display in steps S506 and S518 are changed, and the process proceeds to step S514.

[0221] In step S514, similar to the process in step S502, a request for generating answer information based on the element ID acquired in step S508 is sent to the generation AI server 120, and the process proceeds to step S516.

[0222] In step S516, similar to the processing in step S504, answer information is received from the generation AI server 120, and the process proceeds to step S518, where, similar to the processing in step S304, one of multiple element orders is displayed on the display device 44 based on the multiple element order information contained in the received answer information, and the process proceeds to step S520.

[0223] In step S520, it is determined whether or not the editing by the designer has finished, and if it is determined that the editing has finished (YES), the series of processes is ended.

[0224] On the other hand, if it is determined in step S520 that the editing by the designer has not been completed (NO), the process proceeds to step S508.

[0225] On the other hand, if it is determined in step S510 that the edited result matches the inference result (NO), the process proceeds to step S514.

[0226] [When considering bringing in a piano] Next, the operation when a piano is brought in will be described.

[0227] If a designer wants to install a piano with a width of 120 mm in the plan view of FIG. 9 in CAD software, the designer needs to shorten the width of the toilet to prevent interference between the piano and the toilet wall when bringing the piano in. However, changing the width of the toilet will require editing of other toilet elements as well. Therefore, after changing the width of the toilet, the designer requests inference of the element order to be edited for the toilet. Steps S500 to S506 are then performed to obtain created or edited element IDs from the CAD data currently being edited, and the element order is inferred and displayed from the obtained element IDs. During the inference, the AI ​​model 50 references the learning data in the knowledge base 54. The method for displaying the element order is the same as in the first embodiment.

[0228] Then, when the designer creates or edits the edit element "toilet" based on the inferred and displayed element order, the next element order is inferred and displayed through steps S508 to S518. If the designer's editing result and the inferred result differ, the display rules are changed through step S512.

[0229] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, a request including a generation request to generate multiple element order information pieces that include created or edited element IDs and have different edited elements or editing orders is input to the AI ​​model 50, and multiple element order information pieces are obtained that are output from the AI ​​model 50 in response to the request.

[0230] This makes it possible to grasp a plurality of editing elements that are edited consecutively and the editing order thereof.

[0231] Furthermore, in this embodiment, a request including a generation request to generate multiple pieces of element order information with different editing elements or editing orders is input to the AI ​​model 50, and multiple pieces of element order information output from the AI ​​model 50 are obtained in response to the request, and one of the multiple element orders is displayed based on the obtained multiple pieces of element order information.

[0232] This makes it possible to grasp a plurality of edit elements to be edited successively and a plurality of candidates for the order of editing.

[0233] Furthermore, in this embodiment, learning data including the created or edited element ID, multiple edited elements edited consecutively after that edited element and element order information regarding the order of editing, and evaluation values ​​regarding the editing of the multiple edited elements is registered in the knowledge base 54, and in response to a request, the learning data in the knowledge base 54 is referenced to obtain multiple pieces of element order information output from the AI ​​model 50.

[0234] This allows the order of elements with high evaluation values ​​to be understood. Furthermore, in this embodiment, created or edited element IDs including the element IDs of edited elements created or edited by the designer are obtained based on the displayed element order, and a request including a generation request to generate multiple element order information including the obtained element IDs and with different edit elements or edit orders is input to the AI ​​model 50, and multiple element order information output from the AI ​​model 50 in response to the request is obtained.

[0235] As a result, the element order is inferred and displayed for the results of creation or editing based on the displayed element order, so that a plurality of edit elements that are edited consecutively and their edit order can be grasped.

[0236] Furthermore, in this embodiment, created or edited element IDs including the element IDs of edited elements created or edited by the user are obtained based on the displayed element order, and a request including a generation request to generate multiple element order information including the obtained element IDs and with different edit elements or edit orders is input to the AI ​​model 50, and multiple element order information output from the AI ​​model 50 is obtained in response to the request, and one of the multiple element orders is displayed based on the obtained multiple element order information.

[0237] As a result, a plurality of element orders are inferred and displayed for the results of creation or editing based on the displayed element order, so that a plurality of candidates for the plurality of editing elements to be edited successively and the editing order can be grasped.

[0238] Seventh Embodiment Next, a seventh embodiment of the present invention will be described. Fig. 20 shows this embodiment. Figs. 5, 9, 11 and 18 are also used.

[0239] This embodiment differs from the second and sixth embodiments in that the AI ​​model 50 refers to first learning data including an evaluation value based on a first index or second learning data including an evaluation value based on a second index. Only the differences from the second and sixth embodiments will be described below, and overlapping portions will not be described.

[0240] [Operation of this embodiment] Next, the operation of this embodiment will be described. [Learning data registration process] In the learning data registration process, steps S400 to S402 are performed to obtain the learning data of Figure 5 and the learning data of Figure 11 from the storage device 42, and a request for registration of the learning data is sent to the generation AI server 120. The request includes (1) the learning data of Figure 5 as the first learning data, and (2) the learning data of Figure 11 as the second learning data.

[0241] [Element order information acquisition process] FIG. 20 is a flowchart showing the element order information acquisition process.

[0242] The element order information acquisition process is executed in response to a request from a designer or other user, and when executed by the CPU 30, the process first proceeds to step S530 as shown in FIG.

[0243] In step S530, index information regarding the first index "editing time" or the second index "number of edited items", which indicate an index of the value of the evaluation value, is obtained, and then the process proceeds to step S532, where the created or edited element ID is obtained from the CAD data currently being edited, and the process proceeds to step S534.

[0244] In step S534, a request for generating answer information is sent to the generation AI server 120 based on the index information and element IDs acquired in steps S530 and S532. The request includes (1) the element ID acquired in step S532, (2) element order information relating to multiple edited elements to be edited consecutively after an edited element that has already been created or edited and their edit order, where the multiple element order information has different edited elements or edit orders, (3) a selection request for selecting a predetermined AI model 50, and (4) a reference request for referencing first learning data in the knowledge base 54 if the index related to the index information acquired in step S530 is the first index, or a reference request for referencing second learning data in the knowledge base 54 if the index related to the index information acquired in step S530 is the second index.

[0245] Next, proceed to step S536 to receive answer information from the generation AI server 120, proceed to step S538 to display one of the multiple element orders on the display device 44 based on the multiple element order information contained in the received answer information, similar to the processing of step S304, and proceed to step S540.

[0246] In step S540, if the designer creates or edits an edit element based on the displayed element order, the created or edited element ID including the element ID of the edit element is obtained from the CAD data currently being edited, and the process proceeds to step S542.

[0247] In step S542, based on the element order displayed in steps S538 and S550 and the element ID acquired in step S540, it is determined whether the edited elements created or edited by the designer differ from the edited elements or edit order displayed in steps S538 and S550, and if it is determined that the edited results differ from the inference results (YES), the process proceeds to step S544.

[0248] In step S544, similar to the process in step S310, the display rules used for display in steps S538 and S550 are changed, and the process proceeds to step S546.

[0249] In step S546, similar to the process in step S534, a request for generating answer information based on the index information and element IDs acquired in steps S530 and S540 is sent to the generation AI server 120, and the process proceeds to step S548.

[0250] In step S548, similar to the processing in step S536, answer information is received from the generation AI server 120, and the process proceeds to step S550, where, similar to the processing in step S304, one of multiple element orders is displayed on the display device 44 based on the multiple element order information contained in the received answer information, and the process proceeds to step S552.

[0251] In step S552, it is determined whether or not the editing by the designer has finished, and if it is determined that the editing has finished (YES), the series of processes is ended.

[0252] On the other hand, if it is determined in step S552 that the editing by the designer has not been completed (NO), the process proceeds to step S540.

[0253] On the other hand, if it is determined in step S542 that the edited result matches the inference result (NO), the process proceeds to step S546.

[0254] [When considering bringing in a piano] Next, the operation when a piano is brought in will be described.

[0255] If a designer wants to install a piano with a width of 120 mm in the plan view of Figure 9 in CAD software, he or she needs to shorten the width of the toilet to prevent the piano from interfering with the toilet wall when bringing the piano in. However, changing the width of the toilet will require editing of other elements of the toilet as well. Therefore, after changing the width of the toilet, the designer requests inference of the element order to be edited for the toilet. To obtain an element order that shortens the editing time, the designer selects "editing time" as the indicator. Then, through steps S530 to S538, the created or edited element IDs are obtained from the CAD data currently being edited, and the element order is inferred and displayed from the obtained element IDs. During the inference, the AI ​​model 50 references the first learning data in the knowledge base 54.

[0256] On the other hand, if an element order that reduces the number of edit items is desired, the designer selects "number of edit items" as an index, and steps S530 to S538 are performed to obtain created or edited element IDs from the CAD data currently being edited, and the element order is inferred and displayed from the obtained element IDs. In the inference, the AI ​​model 50 refers to the second learning data in the knowledge base 54.

[0257] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, first learning data including an evaluation value based on the first index and second learning data including an evaluation value based on the second index are registered in the knowledge base 54, index information regarding the first index or the second index is obtained, and a request including a request to refer to either the first learning data or the second learning data in the knowledge base 54 based on the obtained index information is input to the AI ​​model 50.

[0258] This makes it possible to grasp the order of elements with high evaluation values ​​based on the first index or the second index.

[0259] Eighth Embodiment Next, an eighth embodiment of the present invention will be described. This embodiment differs from the sixth and seventh embodiments in that an evaluation value other than the editing time or the number of edited items is adopted. Only the differences from the sixth and seventh embodiments will be described below, and explanations of overlapping parts will be omitted.

[0260] The storage device 42 stores the learning data (hereinafter referred to as "learning data 1") in the third embodiment.

[0261] In the learning data registration process, steps S400 to S402 are performed to obtain learning data 1 from the storage device 42 and send a request to register learning data 1 to the generation AI server 120. When inferring element order information, the generation AI server 120 uses the AI ​​model 50 to refer to learning data 1 in the knowledge base 54, making it possible to determine the element order according to the evaluation value for the quality or risk of the building.

[0262] Ninth Embodiment Next, a ninth embodiment of the present invention will be described. This embodiment differs from the above-mentioned eighth embodiment in that element order information is inferred using multiple pieces of training data each containing evaluation values ​​based on different indices. Below, only the parts that are different from the above-mentioned sixth to eighth embodiments will be described, and a description of the overlapping parts will be omitted.

[0263] [Configuration of this embodiment] First, the configuration of this embodiment will be described. The storage device 42 stores the learning data 1 to 3 in the fourth embodiment.

[0264] In the learning data registration process, steps S400 to S402 are performed to acquire learning data 1 to 3 from the storage device 42, and a request for registering learning data 1 to 3 is sent to the generation AI server 120.

[0265] Steps S502 and S504 consist of first to fourth processes. The first process sends a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. The request includes (1) the element ID acquired in step S500, (2) a generation request to generate multiple pieces of element order information with different edited elements or edit orders, (3) a selection request to select a specific AI model 50, and (4) a reference request to reference the learning data 1 in the knowledge base 54. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0266] The second process sends a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. The request includes (1), (2), and (3) above, as well as (4) a reference request to reference learning data 2 in the knowledge base 54. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0267] The third process is to send a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. In addition to the above (1), (2), and (3), the request also includes (4) a reference request to reference the learning data 3 in the knowledge base 54. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0268] The fourth process determines one or more pieces of element order information from the element order information acquired in the first to third processes. As the fourth process, for example, a process equivalent to the determination unit 466 in the third embodiment can be adopted.

[0269] The same applies to steps S514, S516, steps S534, S536, and steps S546, S548.

[0270] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, element order information output from the AI ​​model 50 is obtained by referencing multiple pieces of learning data each containing evaluation values ​​based on different indices, and the element order information is obtained by determining the element order information from the obtained results.

[0271] This makes it possible to grasp the element order according to the evaluation values ​​based on a plurality of different indices.

[0272] Tenth Embodiment Next, a tenth embodiment of the present invention will be described. This embodiment differs from the above-mentioned eighth embodiment in that element order information is inferred using multiple pieces of training data each containing evaluation values ​​based on different indices. Below, only the parts that are different from the above-mentioned sixth to eighth embodiments will be described, and a description of the overlapping parts will be omitted.

[0273] [Configuration of this embodiment] First, the configuration of this embodiment will be described. The storage device 42 stores the learning data 1 to 3 in the fifth embodiment.

[0274] In the learning data registration process, steps S400 to S402 are performed to acquire learning data 1 to 3 from the storage device 42, and a request for registering learning data 1 to 3 is sent to the generation AI server 120.

[0275] Steps S502 and S504 consist of first to third processes. The first process sends a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. The request includes (1) the element ID acquired in step S500, (2) a generation request to generate multiple pieces of element order information with different edited elements or edit orders, (3) a selection request to select a specific AI model 50, and (4) a reference request to reference the learning data 1 in the knowledge base 54. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0276] The second process sends a request to the generation AI server 120 to generate answer information based on the element order information acquired in the first process. In addition to (2) and (3) above, the request includes (1) the element order information acquired in the first process, and (4) a reference request to reference the learning data 2 in the knowledge base 54. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0277] The third process sends a request to the generation AI server 120 to generate answer information based on the element order information acquired in the second process. In addition to (2) and (3) above, the request includes (1) the element order information acquired in the second process, and (4) a reference request to reference the learning data 3 in the knowledge base 54. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0278] In step S506, one of the plurality of element orders is displayed on the display device 44 based on the element order information acquired in the third process.

[0279] The same applies to steps S514 to S518, steps S534 to S538, and steps S546 to S550.

[0280] [Effects of this embodiment] Next, the effects of this embodiment will be described. In this embodiment, for a request including an element ID and a generation request, element order information output from the AI ​​model 50 is obtained by referencing learning data including an evaluation value based on the first index, and for a request including the obtained element order information and a generation request, element order information output from the AI ​​model 50 is obtained by referencing learning data including an evaluation value based on the second index.

[0281] This makes it possible to grasp the element order according to the evaluation values ​​based on the first index and the second index.

[0282] Furthermore, in this embodiment, for a request including element order information and a generation request, the element order information output from the AI ​​model 50 is obtained by referring to learning data including an evaluation value based on the second index, and for a request including the acquired element order information and a generation request, the element order information output from the AI ​​model 50 is obtained by referring to learning data including an evaluation value based on the third index.

[0283] This makes it possible to grasp the element order according to the evaluation values ​​based on the second index and the third index.

[0284] [Modification] In the first, second, sixth and seventh embodiments and their variations, the editing time or the number of edited items is used as the evaluation value, but this is not limited to this and other evaluation values ​​related to the design of the building can also be used.

[0285] Furthermore, in the third to fifth and eighth to tenth embodiments and their variations, the evaluation values ​​for each index are standardized, but this is not limiting, and normalization or other scaling processing may be performed, or standardization, normalization or other scaling processing may not be performed.

[0286] Furthermore, in the fourth embodiment and its modifications, the quality or risk of the building is used as the index, but the present invention is not limited to this, and user information relating to the user can also be used as the index. For example, the following configuration can be adopted.

[0287] Each row of learning data 1 to 3 includes a user ID for identifying a user (hereinafter referred to as an "editing user") who edited the edited element associated with element ID 410, 430. Explaining using the example of FIG. 5, the second row of FIG. 5 includes two edited elements that were edited consecutively after the edited element in the first row of FIG. 3, the edit order thereof, the user ID of the editing user, and an evaluation value. The third row of FIG. 5 includes two edited elements that were edited consecutively after the edited element in the second row of FIG. 3, the edit order thereof, the user ID of the editing user, and an evaluation value. The fourth row of FIG. 5 includes two edited elements that were edited consecutively after the edited element in the third row of FIG. 3, the edit order thereof, the user ID of the editing user, and an evaluation value.

[0288] In step S302, the user ID of the user currently editing is acquired, and the element ID and user ID acquired in step S300 are input to each of the trained models 460 to 464, and estimation is performed by acquiring the element order information output by the determination unit 466. In this way, element order information suitable for the user can be estimated.

[0289] In this modified example, the editing history data and learning data are configured to include a user ID, but are not limited to this and can be configured to include identification information other than the user ID, or feature information regarding the user's profile, statistics, and other characteristics.

[0290] This modification can be similarly applied to the first to third and fifth to tenth embodiments and their modifications.

[0291] Furthermore, in the above fourth embodiment and its variant, three trained models 460 to 464 are connected in parallel, but this is not limited to this, and the configuration can also be such that two trained models or four or more trained models are connected in parallel.

[0292] Furthermore, in the above fifth embodiment and its variant, three trained models 470 to 474 are connected in series, but this is not limited to this, and the configuration can also be such that two trained models or four or more trained models are connected in series.

[0293] Furthermore, in the above ninth and tenth embodiments and their variations, the AI ​​model 50 is made to refer to three pieces of learning data, but this is not limited to this, and it is also possible to make the AI ​​model refer to two pieces of learning data or four or more pieces of learning data.

[0294] In addition, in the fourth and fifth embodiments and their modifications, the plurality of trained models are configured by connecting them in parallel or in series, but this is not limiting, and any combination of parallel and series connections can be used. In a node that aggregates the output results of the plurality of trained models, a configuration including a determination unit 466 can be adopted when determining element order information from the output results of the plurality of trained models, or a configuration in which the output results of the plurality of trained models are input directly to the trained model of the next stage.

[0295] In the fourth embodiment and its modifications, the determination unit 466 determines one or more pieces of element order information from the element order information output by the trained models 460 to 464. However, the present invention is not limited to this, and any method can be adopted as a method for determining element order information. The same applies to the fourth process in the ninth embodiment.

[0296] Furthermore, in the first to fifth embodiments and their modifications, a plurality of pieces of element order information are estimated in steps S302, S312, S336, and S346, but this is not limiting, and one piece of element order information can also be estimated.

[0297] Furthermore, in the sixth to tenth embodiments and their variations, in steps S502, S514, S534, and S546, a generation request for generating multiple pieces of element order information is included in the request, but this is not limited to this, and a generation request for generating one piece of element order information can also be included in the request.

[0298] Furthermore, in the above first to tenth embodiments and their variations, display rules 1 to 4 were adopted, but this is not limited to this. For example, a display rule that displays the fewest number of editing elements, an element order above a predetermined number, below a predetermined number, or a predetermined order, or any other display rule can be adopted.

[0299] Furthermore, in the above first to tenth embodiments and their variations, the element order is displayed according to one of a plurality of display rules, but this is not limited to this, and the element order can be displayed according to a predetermined display rule without providing a plurality of display rules.

[0300] Furthermore, in the first to tenth embodiments and their modifications, one element order is displayed, but the present invention is not limited to this and multiple element orders can be displayed.

[0301] Furthermore, in the first to tenth embodiments and their modifications, the display rules are changed, but the present invention is not limited to this, and it is also possible to employ a configuration in which the display rules are not changed.

[0302] Furthermore, in the first to fifth embodiments and their modifications, estimation and display are repeated multiple times, but this is not limiting, and a configuration in which estimation and display are performed only once can be adopted. The same applies to the sixth to tenth embodiments and their modifications.

[0303] Furthermore, in the above first to fifth embodiments and their variations, reinforcement learning is adopted as the learning method, but this is not limited to this, and any learning method such as supervised learning, semi-supervised learning, unsupervised learning, deep learning, or the like can be adopted.

[0304] Furthermore, in the first to tenth embodiments and their modifications, the learning data includes evaluation values, but this is not limiting, and the learning data may be configured without including evaluation values.

[0305] Furthermore, in the above first to tenth embodiments and their variations, the editing history data and learning data are configured to include element IDs, but are not limited to this and may be configured to include identification information other than element IDs, or feature information regarding summaries, statistics, and other features of the editing elements.

[0306] In the second embodiment and its modifications, the first trained model and the second trained model can be configured as one trained model. Similarly, in the third to fifth embodiments and their modifications, m (m≧2) trained models can be configured as n (1≦n≦m) trained models.

[0307] In the sixth to tenth embodiments and their modifications, the AI ​​model 50 can be configured as the trained model in the first to fifth embodiments and their modifications.

[0308] Furthermore, in the above first to tenth embodiments and their variations, an element order that appears a predetermined number of times or more in the learning data is identified, but this is not limited to this, and an element order that appears a predetermined number of times or more in the editing history data can also be identified.

[0309] Furthermore, in the first to tenth embodiments and their modifications, a maximum of four editing elements and their editing orders are handled, but this is not limiting, and five or more editing elements and their editing orders can be handled. There is no need to limit the number of editing elements as long as there are multiple editing elements. Furthermore, the number of editing elements can be set independently for each of the learning data in FIGS. 5 and 11.

[0310] Furthermore, in the above-described first to tenth embodiments and their modifications, all combinations of element orders with two to four edited elements are obtained from the edit history data to generate learning data, but this is not limiting. It is also possible to obtain element orders that appear a predetermined number of times or more in the edit history data from the edit history data to generate learning data. In this case, an element order with a high frequency of appearance may be estimated using a trained model trained on multiple edit history data. This allows the element order with a high frequency of appearance in the edit history data to be learned or inferred.

[0311] Furthermore, in the first to tenth embodiments and their modifications, the edit history data is configured as data separate from the CAD data, but this is not limiting, and the edit history data can be configured as an integral part of the CAD data.

[0312] Furthermore, in the sixth to tenth embodiments and their modifications, the AI ​​model 50 is made to refer to the learning data in the knowledge base 54, but this is not limiting, and the learning data to be referred to in the knowledge base 54 can be included in the request. This makes it possible to apply the present invention to a configuration that does not include the knowledge base 54. Specifically, for example, the following configuration can be adopted.

[0313] [Invention A1] An element information acquisition means for acquiring element information relating to created or edited elements in design information; an input means for inputting a request to the AI ​​model, the request including the element information acquired by the element information acquisition means, element information relating to an element that has already been created or edited, element order information relating to a plurality of elements that have been edited consecutively after the element and the order of editing thereof (hereinafter, "a plurality of elements and the order of editing thereof" will be referred to as "element order"), and reference information including evaluation values ​​relating to the plurality of elements, and a request for generating element order information relating to the element order; an acquisition means for acquiring the element order information output from the AI ​​model in response to the request, The evaluation value is an evaluation value relating to the quality or risk of the design object.

[0314] This makes it possible to grasp the plurality of elements that are edited consecutively and the order in which they are edited, as well as the order in which elements are edited according to the evaluation values ​​regarding the quality or risk of the object of design.

[0315] [Invention A2] In Invention A1, an index information acquisition means for acquiring index information relating to a first index indicating an index of the value of the evaluation value or a second index different from the first index; The input means inputs the request to the AI ​​model based on the index information acquired by the index information acquisition means, the request including either first reference information including the element information, the element order information, and the evaluation value based on the first index, or second reference information including the element information, the element order information, and the evaluation value based on the second index.

[0316] This makes it possible to grasp the order of elements with high evaluation values ​​based on the first index or the second index.

[0317] [Invention A3] In Inventions A1 and A2, The evaluation value regarding the quality of the design object is an evaluation value regarding the safety, functionality, convenience, aesthetics, environmental friendliness, economic efficiency, or legal compliance of the design object.

[0318] [Invention A4] In Inventions A1 and A2, The evaluation value regarding the risk of the design object is an evaluation value regarding the risk of a decline in the quality of the design object, a risk of the design object collapsing or being damaged, a risk of the design object deteriorating, a risk of an increase in costs related to the design object, a risk to the design object due to a natural disaster, a risk of a decrease in the asset value related to the design object, or a risk of the design object causing a social problem.

[0319] As modifications of the ninth embodiment, embodiments A1 to A4 will be described. The first process sends a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. The request includes (1) the element ID acquired in step S500, (2) a generation request to generate multiple pieces of element order information with different edited elements or edit orders, (3) a selection request to select a specific AI model 50, and (4) training data 1. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0320] The second process sends a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. The request includes (1), (2), and (3) above, as well as (4) learning data 2. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0321] The third process sends a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. The request includes (1), (2), and (3) above, as well as (4) learning data 3. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0322] The fourth process determines one or more pieces of element order information from the element order information acquired in the first to third processes. As the fourth process, for example, a process equivalent to the determination unit 466 in the third embodiment can be adopted.

[0323] The same applies to steps S514, S516, steps S534, S536, and steps S546, S548.

[0324] Next, embodiments of inventions A1 to A4 will be described as modifications of the tenth embodiment.

[0325] The first process sends a request to the generation AI server 120 to generate answer information based on the element ID acquired in step S500. The request includes (1) the element ID acquired in step S500, (2) a generation request to generate multiple pieces of element order information with different edited elements or edit orders, (3) a selection request to select a specific AI model 50, and (4) training data 1. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0326] The second process sends a request to the generation AI server 120 to generate answer information based on the element order information acquired in the first process. In addition to (2) and (3) above, the request includes (1) the element order information acquired in the first process and (4) learning data 2. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0327] The third process sends a request to the generation AI server 120 to generate answer information based on the element order information acquired in the second process. In addition to (2) and (3) above, the request includes (1) the element order information acquired in the second process and (4) learning data 3. Then, answer information is received from the generation AI server 120 in response to the request, and multiple pieces of element order information are acquired from the received answer information.

[0328] The same applies to steps S514 to S518, steps S534 to S538, and steps S546 to S550.

[0329] [Invention A5] In Invention A1, The plurality of elements are elements that require human judgment to set or change, and the setting or change affects other elements.

[0330] [Invention A6] In Invention A1, The design information is information for designing a building.

[0331] Furthermore, in the sixth to tenth embodiments and their modifications, the AI ​​model 50 references information in the knowledge base 54, but the present invention is not limited to this. Information to be referenced in the knowledge base 54 (for example, the learning data in FIG. 5 or FIG. 11) can be obtained by a web search or the like, and the search results can be referenced by the AI ​​model 50 to perform inference. This configuration can be realized, for example, by RAG (Retrieval Augmented Generation).

[0332] Furthermore, in the first to tenth embodiments and their modifications, the trained model or AI model 50 is used, but the present invention is not limited to this, and for example, the following configuration can be adopted.

[0333] [Invention B1] An element information acquisition means for acquiring element information relating to created or edited elements in design information; a search means for searching, from a storage means for storing element order information relating to a plurality of elements edited consecutively after an element that has already been created or edited and the order in which they are edited (hereinafter, "a plurality of elements and the order in which they are edited" will be referred to as "element order") in association with the element information relating to the created or edited elements and evaluation values ​​relating to the plurality of elements, for element order information corresponding to the element information acquired by the element information acquisition means and having an evaluation value equal to or greater than a predetermined value; The evaluation value is an evaluation value relating to the quality or risk of the design object.

[0334] This makes it possible to grasp the plurality of elements that are edited consecutively and the order in which they are edited, as well as the order in which elements are edited according to the evaluation values ​​regarding the quality or risk of the object of design.

[0335] [Invention B2] In Invention B1, The evaluation value regarding the quality of the design object is an evaluation value regarding the safety, functionality, convenience, aesthetics, environmental friendliness, economic efficiency, or legal compliance of the design object.

[0336] [Invention B3] In Invention B1, The evaluation value regarding the risk of the design object is an evaluation value regarding the risk of a decline in the quality of the design object, a risk of the design object collapsing or being damaged, a risk of the design object deteriorating, a risk of an increase in costs related to the design object, a risk to the design object due to a natural disaster, a risk of a decrease in the asset value related to the design object, or a risk of the design object causing a social problem.

[0337] As modifications of the fourth embodiment, embodiments of inventions B1 to B3 will be described. The storage device 42 stores element order information tables 1 to 3 having the same data structure as the learning data 1 to 3 in the fourth embodiment. In step S302, element order information tables 1 to 3 are searched for one or more pieces of element order information corresponding to the element ID acquired in step S300 and having an evaluation value equal to or greater than a predetermined value, respectively, and one or more pieces of element order information are determined from the retrieved element order information. This determination method may be the same as the determination method used by the determination unit 466. In step S304, the element order is displayed on the display device 44 based on the element order information determined in step S302. The same applies to steps S312, S134, S336, S338, S346, and S348.

[0338] Next, embodiments of inventions B1 to B3 will be described as modifications of the fifth embodiment. The storage device 42 stores element order information tables 1 to 3, each having the same data structure as the learning data 1 to 3 in the fifth embodiment. In step S302, one or more pieces of element order information corresponding to the element ID acquired in step S300 and having an evaluation value equal to or greater than a predetermined value are acquired from element order information table 1. Next, one or more pieces of element order information corresponding to the acquired element order information and having an evaluation value equal to or greater than a predetermined value are acquired from element order information table 2. Then, one or more pieces of element order information corresponding to the acquired element order information and having an evaluation value equal to or greater than a predetermined value are acquired from element order information table 3. In step S304, the element order is displayed on the display device 44 based on the element order information acquired in step S302. The same applies to steps S312, S314, S336, S338, S346, and S348.

[0339] [Invention B4] In Inventions B1 to B3, the storage means includes a first storage means for storing the element order information in association with the element information and the evaluation value based on a first index indicating an index of value of the evaluation value, and a second storage means for storing the element order information in association with the element information and the evaluation value based on a second index different from the first index; index information acquisition means for acquiring index information relating to the first index or the second index; a storage means selection means for selecting either the first storage means or the second storage means based on the index information acquired by the index information acquisition means, The retrieval means retrieves the element order information from the storage means selected by the storage means selection means.

[0340] This makes it possible to grasp the order of elements with high evaluation values ​​based on the first index or the second index.

[0341] An embodiment of invention B4 will be described as a modification of the fourth embodiment. The storage device 42 stores element order information tables 1 and 2, each having the same data structure as the learning data 1 and 2 in the fourth embodiment. In step S332, element order information table 1 is selected if the index related to the index information acquired in step S330 is the first index, and element order information table 2 is selected if the index related to the acquired index information is the second index. In step S336, one or more pieces of element order information corresponding to the element ID acquired in step S334 and having an evaluation value equal to or greater than a predetermined value are searched from the element order information table selected in step S332. In step S346, one or more pieces of element order information corresponding to the element ID acquired in step S340 and having an evaluation value equal to or greater than a predetermined value are searched from the element order information table selected in step S332.

[0342] [Invention B5] In Inventions B1 to B3, the storage means stores the element order information in association with the element information, the evaluation value, and index information relating to a first index indicating an index of value of the evaluation value or a second index different from the first index; index information acquisition means for acquiring index information relating to the first index or the second index; The search means searches for the element order information corresponding to the element information acquired by the element information acquisition means and the index information acquired by the index information acquisition means.

[0343] This makes it possible to grasp the order of elements with high evaluation values ​​based on the first index or the second index.

[0344] An embodiment of invention B5 will be described as a modification of the fourth embodiment. The storage device 42 stores an element order information table in which, for each row, created or edited element ID 430, element order information 432, evaluation values ​​414, 434, and index information related to the first index or the second index are registered. In step S336, the element order information table is searched for one or more pieces of element order information corresponding to the element ID and index information acquired in steps S330 and S334, and having an evaluation value equal to or greater than a predetermined value. In step S346, the element order information table is searched for one or more pieces of element order information corresponding to the element ID and index information acquired in steps S330 and S340, and having an evaluation value equal to or greater than a predetermined value.

[0345] [Invention B6] In Invention B1, The plurality of elements are elements that require human judgment to set or change, and the setting or change affects other elements.

[0346] [Invention B7] In Invention B1, The design information is information for designing a building.

[0347] Furthermore, in inventions B1 to B7 and their variations, storing element order information in association with element information, etc. includes, for example, (1) storing element order information and element information, etc. in the same record in a direct association manner, or (2) storing element order information via one or more intermediate pieces of information, such as providing a table in which element order information and intermediate information are associated and registered, and another table in which element information, etc. and intermediate information are associated and registered. In other words, any data structure can be adopted as long as it is possible to trace element order information from element information, etc. Note that element order information may be stored in a storage means in association with element information, etc., and it is not necessarily required that element information, etc., be stored in the storage means.

[0348] Furthermore, in inventions B1 to B7 and their variations, the storage means stores the element order information by any means and at any time, and may store the element order information in advance, or may store the element order information by external input or the like during operation of the drawing creation support device 100 without storing the element order information in advance.

[0349] Furthermore, in the sixth to tenth embodiments and their variations, the prompt is, for example, a request to the AI ​​model 50 to generate, based on the created or edited element ID, a plurality of edited elements to be edited successively after the created or edited edited element and element order information relating to the edit order, which has an evaluation value of a predetermined value or more. However, the prompt is not limited to this, and the prompt can be a request to the AI ​​model 50 to generate a plurality of edited elements to be edited successively after the created or edited edited element and element order information relating to the edit order.

[0350] 5 or 11 is registered in the knowledge base 54 in the sixth to tenth embodiments and their modifications. However, the present invention is not limited to this. (1) Edit history data; (2) Reference information including element information on an element that has already been created or edited, and element order information on multiple elements to be edited in succession after the element and the order in which they are edited; or (3) Reference information including element information on an element that has already been created or edited, element order information on multiple elements to be edited in succession after the element and the order in which they are edited, and evaluation values ​​related to the editing of the multiple elements. The reference information in (2) or (3) may include element order information estimated using the trained model in the first to fifth embodiments and their modifications.

[0351] Furthermore, in the sixth to tenth embodiments and their modifications, vector data is registered in the knowledge base 54, but this is not limiting and data in any format can be registered.

[0352] Furthermore, in the first to tenth embodiments and their modifications, a configuration including any one of an estimation process using a trained model, an acquisition process from the AI ​​model 50, and a search process using a table is employed, but the present invention is not limited to this, and a configuration including two or more of these processes may be employed. Specifically, for example, the following configuration may be employed.

[0353] In the first configuration, the process of step S302 or the process of step S312 is performed by any one of estimation process, acquisition process, and search process. The process of step S312 from the second time onwards can be performed by the same or different process as the process of step S302 or the first time.

[0354] In the second configuration, the process of step S336 or the process of step S346 is performed by any one of estimation process, acquisition process, and search process. The process of step S346 from the second time onwards can be performed by the same or different process as the process of step S336 or the first time.

[0355] In the third configuration, the processing of steps S502 and S504 or the processing of steps S514 and S516 is performed by any of estimation processing, acquisition processing, and search processing. The second and subsequent processing of steps S514 and S516 can be performed by the same or different processing as the processing of steps S502 and S504 or the first processing of steps S514 and S516.

[0356] In the fourth configuration, the processing of steps S534 and S536 or the processing of steps S546 and S548 is performed by any one of estimation processing, acquisition processing, and search processing. The second and subsequent processing of steps S546 and S548 can be performed by the same or different processing as the first processing of steps S534 and S536 or steps S546 and S548.

[0357] The fifth configuration is a configuration for controlling which process is given priority. For example, (1) a configuration in which a process with a low current load is given priority among a plurality of processes, (2) a configuration in which a process with a high degree of adoption (number of times adopted, percentage, or other degree) by users is given priority among a plurality of processes, or (3) a configuration in which a process with a high degree of use (number of times used, percentage, or other degree) by users is given priority among a plurality of processes can be adopted.

[0358] Furthermore, in the first to tenth embodiments and their modifications, multiple edit elements and their edit order are learned or inferred, but this is not limiting. The "multiple edit elements" to be learned or inferred can be edit element A, whose setting or change requires human judgment, and whose setting or change affects another edit element B. This makes it possible to understand the element order taking into account the relationship between edit element A and edit element B. For edit elements that do not require human judgment, editing can be automated using the technology of Japanese Patent No. 7341580, for example. Therefore, by targeting edit elements that are difficult to automate, the editing work can be made more efficient.

[0359] In the first to fifth embodiments and their modifications, the device is implemented as a single device, but the present invention is not limited to this and can also be implemented as a network system. As an example of a network system, some or all of the functions of the drawing creation support device 100 can be configured as a virtual server on a server that provides a cloud computing service.

[0360] Furthermore, in the sixth to tenth embodiments and their variations, the generation AI server 120 is configured as an integrated unit that includes the functions of the AI ​​model 50, the AI ​​model control unit 52, the knowledge base 54, the request receiving unit 56, the request processing unit 58, the answer information sending unit 60, the request receiving unit 62, and the learning data registration unit 64, but this is not limited to this, and some of the functions can be configured as separate servers, etc.

[0361] Furthermore, in the sixth to tenth embodiments and their modifications, the system is realized as a network system, but the present invention is not limited to this and can be realized as a single device or application.

[0362] Furthermore, in the sixth to tenth embodiments and their modifications, the case where the present invention is applied to a network system consisting of the Internet 199 has been described, but the present invention is not limited to this and may be applied to, for example, a so-called intranet that communicates in the same manner as the Internet 199. Of course, the present invention is not limited to a network that communicates in the same manner as the Internet 199, but can be applied to a network of any communication method.

[0363] Furthermore, in the above first to tenth embodiments and their variations, the drawing creation support device 100 is configured to use the storage device 42, but this is not limited to this, and it can also be configured to use an external storage device such as a database server.

[0364] Furthermore, in the above first to tenth embodiments and their variations, when executing the processes shown in the flowcharts of Figures 4, 6, 8, 12, and 18 to 20, we have described the case where a program pre-stored in ROM 32 is executed, but this is not limited to this, and the program showing these procedures may be read into RAM 34 from a storage medium on which the program is stored and executed.

[0365] Here, storage media refers to semiconductor storage media such as RAM and ROM, magnetic storage media such as FD and HD, optically readable storage media such as CD, CDV, LD and DVD, and magnetic storage / optically readable storage media such as MO, and includes all storage media that can be read by a computer, regardless of the reading method (electronic, magnetic, optical, etc.).

[0366] Moreover, the first to tenth embodiments and their modifications can be applied to each other. Furthermore, the present invention is not limited to the first to tenth embodiments and their modifications, and can be applied to other cases without departing from the spirit of the present invention. For example, the present invention can be applied to a wide range of design tasks, such as automobile design, machine design, and circuit design (when the design targets are automobiles, machines, circuits, and the like). [Explanation of symbols]

[0367] 100...Drawing creation support device, 30...CPU, 32...ROM, 34...RAM, 38...I / F, 39...bus, 40...input device, 42...storage device, 44...display device, 120...generation AI server, 50...AI model, 52...AI model control unit, 54...knowledge base, 56, 62...request receiving unit, 58...request processing unit, 60...answer information sending unit, 64...learning data registration unit, 199...Internet, 400, 420...element information, 402...editing time, 422...number of edited items, 410, 430...element ID, 412, 432...element order information, 414, 434...evaluation value, 450-454...CAD data, 460-464, 470-474...trained model, 466...determination unit

Claims

1. an element information acquisition means for acquiring element information relating to created or edited elements in the design information; an estimation means for estimating the element order information from the element information acquired by the element information acquisition means, using a trained model trained on training data including element information on an element that has been created or edited, element order information on a plurality of elements that have been edited consecutively after the element and the order of editing the elements (hereinafter, "the plurality of elements and the order of editing the elements" will be referred to as "element order"), and evaluation values ​​on the plurality of elements; A design support system, wherein the evaluation value is an evaluation value relating to the quality or risk of a design object.

2. In claim 1, A design support system characterized in that the evaluation value regarding the quality of the design object is an evaluation value regarding the safety, functionality, convenience, aesthetics, environmental friendliness, economic efficiency, or legal compliance of the design object.

3. In claim 1, A design support system characterized in that the evaluation value regarding the risk of the design object is an evaluation value regarding the risk of a decline in the quality of the design object, a risk of the design object collapsing or being damaged, a risk of the design object deteriorating, a risk of an increase in costs related to the design object, a risk to the design object due to a natural disaster, a risk of a decrease in asset value related to the design object, or a risk of the design object causing a social problem.

4. In any one of claims 1 to 3, A design support system characterized in that the trained model is trained to maximize the evaluation value based on training data including the element information, the element order information, and the evaluation value.

5. In claim 4, The trained model includes: a first trained model trained to maximize an evaluation value based on training data including the element information, the element order information, and an evaluation value based on a first index indicating an index of value of the evaluation value; and a second trained model trained to maximize an evaluation value based on training data including the element information, the element order information, and an evaluation value based on a second index different from the first index, index information acquisition means for acquiring index information relating to the first index or the second index; a trained model selection means for selecting either the first trained model or the second trained model based on the index information acquired by the index information acquisition means; A design support system characterized in that the estimation means estimates the element order information using the trained model selected by the trained model selection means.

6. In claim 1, A design support system characterized in that the plurality of elements are elements whose setting or change requires human judgment and whose setting or change has an effect on other elements.

7. In claim 1, A design support system, wherein the design information is design information for designing a building.

8. a calculation means for calculating an evaluation value of the element order information based on first design information including element order information relating to a plurality of elements that have been edited consecutively and their edit order (hereinafter, "the plurality of elements and their edit order" will be referred to as "element order") and having a first evaluation value set thereto, and second design information including element order information relating to the element order and having a second evaluation value set thereto; generating means for generating learning data including the element order information included in the first design information and the second design information and the evaluation value calculated by the calculation means.

9. In claim 8, The calculation means calculating an evaluation value of the element order information included in the first design information based on the first evaluation value; calculating an evaluation value of the element order information included in the second design information based on the second evaluation value; a learning data generation system that calculates an evaluation value of the element order information that is commonly included in the first design information and the second design information, based on the first evaluation value and the second evaluation value.

10. generation means for generating a trained model by performing training based on training data to which evaluation values ​​are assigned for a combination of element information on an element that has already been created or edited and element order information on a plurality of elements that have been edited consecutively after the element and the order in which they are edited (hereinafter, "the plurality of elements and the order in which they are edited" will be referred to as "element order"), so that the evaluation value is maximized; The trained model generation system, wherein the evaluation value is an evaluation value regarding the quality or risk of the design object.

Images