Analysis method, classification method, classification device, estimation method, and estimation device

Abstract

To provide a novel method for analysis, a method for classification, a classification device, a method for estimation, and an estimation device about the relation between a subjective evaluation and an evaluation condition.SOLUTION: The present invention relates to a method for specifying a subjective evaluation which changes according to conditions, the method including the steps of: acquiring an evaluation value based on a subjective view about a target of a subject with an evaluation term as an index under different conditions; acquiring condition information showing a condition at the time of an evaluation; and performing a correlation analysis of specific condition information and the evaluation value for each of a plurality of evaluation terms and specifying an evaluation term that has a correlation with the condition information.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an analysis method or apparatus for analyzing changes in subjective evaluations of humans under various conditions, a classification method or apparatus for subjects, an estimation method or apparatus for subjective evaluations, and an estimation method or apparatus for the state of a subject.

Background Art

[0002] Subjective evaluations are generally obtained by methods such as questionnaires, and human subjectivity is rated using a five-point scale evaluation or the like. It is mainly used in surveys on human behavior and awareness, and surveys on satisfaction and purchase intention of products.

[0003] It is known that subjective evaluations include evaluations of awareness and behavior for hearing consumers' daily awareness and behavior, and evaluations of feelings about events and objects in the environment. Among them, the ability to identify feelings towards stimuli from events and objects in the environment is sensitivity.

[0004] Generally, it is empirically recognized that even when the target event (object) itself does not change, such as feeling that food tastes good when in a good mood, the sensitivity to that object changes under various conditions.

[0005] As a technique for estimating emotions, for example, Patent Document 1 describes estimating a user's emotion based on the user's voice. However, there has been no conventional method for scientifically analyzing the relationship between various conditions including the emotions of subjects and sensitivity.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0007] The present invention aims to clarify the relationship between subjective evaluations of objects and the conditions under which those evaluations are conducted, by providing a novel analytical method for the relationship between subjective evaluations and evaluation conditions. [Means for solving the problem]

[0008] To solve the above problems, the present invention provides a method for identifying subjective evaluations that vary depending on conditions, The steps include obtaining subjective evaluation values ​​from subjects, using evaluation terms as indicators, for the object under multiple different conditions, and Steps include obtaining conditional information that indicates the conditions during evaluation, The method includes the step of performing a correlation analysis between specific condition information and the evaluation value for each of several evaluation terms, and identifying evaluation terms that have a correlation with the condition information.

[0009] This configuration allows for the objective identification of evaluation terms whose evaluation values ​​change depending on variations in specific conditions for a given object.

[0010] In a preferred embodiment of the present invention, the evaluation value is a sensitivity evaluation value indicating the strength with which a subject feels the sensation represented by the evaluation term.

[0011] In a preferred embodiment of the present invention, the conditions include environmental conditions that describe the environment surrounding the subject.

[0012] In a preferred embodiment of the present invention, the environmental conditions include a time period within a day.

[0013] In a preferred embodiment of the present invention, the conditions include state conditions that indicate the state of the subject.

[0014] In a preferred embodiment of the present invention, the state conditions include the age or age group of the subject.

[0015] In a preferred embodiment of the present invention, the state conditions include the mental state of the subject.

[0016] In a preferred embodiment of the present invention, the condition information indicating the state conditions includes emotional valence indicating the subject's emotions.

[0017] In a preferred embodiment of the present invention, the condition information indicating the state conditions includes the subject's level of alertness.

[0018] To solve the above problems, the present invention provides a method for identifying conditions that influence subjective evaluation, comprising the steps of: obtaining subjective evaluation values ​​of an object, using evaluation terms as indicators, under a plurality of different conditions; obtaining a plurality of types of condition information indicating the conditions at the time of evaluation; and performing a correlation analysis between the evaluation value and the condition information for a specific evaluation term for each of the plurality of types of condition information, thereby identifying the condition information that has a correlation with the evaluation value.

[0019] This configuration allows for the objective identification of conditions that influence fluctuations in evaluation values, which are based on specific evaluation terms, for a given object.

[0020] To solve the above problems, the present invention provides an analytical method for analyzing the properties of an object that cause changes in subjective evaluations of the object due to changes in specific conditions, comprising: determining an evaluation term as a specific term for an object whose evaluation value changes under specific conditions; and identifying the properties of the object that affect the evaluation value of the specific term, wherein the step of determining the specific term includes: obtaining subjective evaluation values ​​of an object, using the evaluation term as an indicator, under multiple different conditions; obtaining condition information indicating the conditions at the time of evaluation; performing a correlation analysis between the specific condition information and the evaluation value for each of the multiple evaluation terms, and determining the evaluation term for which the evaluation value has a correlation relationship with the condition information as the specific term, wherein the step of identifying the properties of the object includes: varying the condition information for multiple objects with different properties, obtaining evaluation values ​​using the specific term as an indicator for each object and condition information, respectively; and identifying properties that affect the evaluation value of the specific term based on the properties of the multiple objects and the changes in the evaluation value according to the condition information.

[0021] This configuration allows us to identify evaluation terms whose evaluation values ​​fluctuate under specific conditions, and to identify the properties of objects that influence the fluctuations of those evaluation terms under those conditions. This means that, for example, by designing a product with those properties as the target, we can design a product whose evaluation values ​​fluctuate under specific conditions.

[0022] In a preferred embodiment of the present invention, the step of identifying the properties of an object includes: varying the condition information for a plurality of objects of the same type as one object in the step of determining the specific term, but with different properties, and obtaining evaluation values ​​using the specific term as an indicator for each object and condition information; identifying a plurality of objects from among the plurality of objects in which the evaluation value changes to a threshold or greater in response to the variation in the condition information; and identifying a property common to the plurality of identified objects as a property that affects the specific term.

[0023] In a preferred embodiment of the present invention, the step of specifying the property that affects the evaluation value of the specific term includes analyzing the relationship between the properties of a plurality of objects with different properties and the relationship information indicating the relationship between the condition information and the evaluation value in each object, and specifying the property that affects the evaluation value of the specific term based on the relationship.

[0024] In a preferred embodiment of the present invention, the property includes a value representing the mechanical property of the object.

[0025] In a preferred embodiment of the present invention, the property includes a value representing the chemical property of the object.

[0026] In order to solve the above problems, the present invention is a classification method for classifying the variation characteristics of the subjective evaluation of a subject, including the steps of obtaining the subjective evaluation value of the subject using an evaluation term as an index for an object under specific conditions, obtaining condition information indicating the conditions at the time of evaluation, and specifying the classification of the subject based on the variation characteristics of the evaluation value according to the condition information by combining the evaluation value and the condition information.

[0027] With such a configuration, classification according to individual differences in the variation of the "feeling" evaluated by the evaluation term becomes possible.

[0028] In a preferred embodiment of the present invention, the method further includes storing recommended products based on the classification in a database and determining the recommended products based on the classification of the subject according to the variation characteristics by referring to the database. With such a configuration, it becomes possible to determine recommended products based on the classification of the subject and make a more satisfactory proposal considering the variation of the evaluation due to changes in conditions.

[0029] To solve the above problems, the present invention provides a classification device for classifying the variability characteristics of subjective evaluations by subjects, comprising: means for acquiring subjective evaluation values ​​of an object, using evaluation terms as indicators, under specific conditions; means for acquiring condition information indicating the conditions at the time of evaluation; and means for identifying the classification of subjects based on the variability characteristics of the evaluation values ​​according to the condition information, based on a combination of the evaluation values ​​and the condition information.

[0030] To solve the above problems, the present invention is an estimation method that estimates a subject's subjective evaluation value of an object, using specific evaluation terms as indicators, based on condition information indicating the conditions at the time of evaluation.

[0031] In a preferred embodiment of the present invention, the proposal to the subject is further determined based on the estimation results of the evaluation values.

[0032] This configuration allows us to estimate the subjective evaluation values ​​of the subjects from the given conditions, without requiring them to directly evaluate the object. This enables us to make more appropriate suggestions based on the estimated results.

[0033] In a preferred embodiment of the present invention, the evaluation value is estimated based on property information indicating the properties of the object, in addition to the condition information.

[0034] To solve the above problems, the present invention is an estimation device that estimates a subject's subjective evaluation value of an object, using specific evaluation terms as indicators, based on condition information indicating the conditions at the time of evaluation.

[0035] To solve the above problems, the present invention is an estimation method that estimates the state of a subject at the time of evaluation based on subjective evaluation values ​​of an object, using evaluation terms as indicators.

[0036] This configuration allows for the estimation of the subject's state based on the evaluation of the object using evaluation terms as indicators. This enables more appropriate suggestions and other actions based on the estimation results.

[0037] In a preferred embodiment of the present invention, the state of the subject is estimated based on property information indicating the properties of the object, in addition to the evaluation value.

[0038] To solve the above problems, the present invention is an estimation device that estimates the state of a subject at the time of evaluation based on the subject's subjective evaluation value, which is based on evaluation terms used as indicators for an object. [Effects of the Invention]

[0039] According to the present invention, a novel analytical method concerning the relationship between subjective evaluation and evaluation conditions can be provided. [Brief explanation of the drawing]

[0040] [Figure 1] This is a flowchart relating to the correlation analysis of this embodiment. [Figure 2] This figure shows an example of the information obtained in the correlation analysis of this embodiment. [Figure 3] This is a flowchart relating to the design method of this embodiment. [Figure 4] This is a functional block diagram relating to the classification device of this embodiment. [Figure 5] This is a flowchart relating to the classification method of this embodiment. [Figure 6] This is a functional block diagram relating to the estimation device of this embodiment. [Figure 7] This is a flowchart relating to the estimation method of this embodiment. [Figure 8] This figure shows the combinations that showed a correlation between the correlation coefficient between emotional valence and evaluation value and the physical properties of the object in the example. [Figure 9] This figure shows the combinations of the number of subjects who showed a correlation between emotional valence and evaluation value, and the physical properties of the object, in the example. [Modes for carrying out the invention]

[0041] <Definition> This invention relates to a technique for analyzing, estimating, or classifying the correlation between evaluation values, which are indexed by evaluation terms, and evaluation conditions. In this invention, evaluation is performed on a specific object by subjective evaluation by a subject based on specified evaluation terms. In this invention, evaluation terms refer to words that express subjective evaluations. In this embodiment, for example, these include terms that express emotions (emotional terms) such as like, dislike, fun, sad, and surprise; terms that express abstract impressions of an object (impressional terms) such as luxury, attractive, special, and gentle; and terms that express sensations directly evoked by the object itself (sensory terms) such as hard, soft, moist, and dry. Of these, emotional terms and impressional terms are called "higher-order terms," ​​and sensory terms are called "lower-order terms."

[0042] Furthermore, evaluation values ​​based on evaluation terms refer to values ​​obtained through subjective evaluations by subjects, based on how well the object fits the impression represented by each evaluation term. Preferably, the evaluation value is a sensitivity evaluation value indicating the strength with which the subject feels the sensation represented by the evaluation term. In this embodiment, the evaluation value is expressed as an integer between 0 and 100, but the minimum and maximum values ​​can be arbitrarily defined. Furthermore, the subject's subjective evaluation may be obtained as a binary value of whether or not the evaluation term for the object is appropriate, and this may also be used as the evaluation value.

[0043] In this invention, a subject evaluates an object under various conditions, obtains conditional information indicating the conditions in the evaluation, and performs a correlation analysis between the evaluation value and the conditional information. The conditions include environmental conditions relating to the environment surrounding the subject at the time of evaluation and state conditions relating to the subject's state, and at least one of these is obtained. Environmental conditions include, for example, the month, time, time of day, temperature, humidity, and climate in which the evaluation was performed, and information indicating these is obtained as conditional information. State conditions include biological attributes such as the subject's age and gender, and the subject's mental state, and information indicating these is obtained as conditional information.

[0044] Here, regarding the conditional information, conditions such as temperature, humidity, the subject's age, and mental state can be expressed numerically (numerical conditions), while environmental conditions such as time of day and state conditions such as gender are expressed as choices rather than numerically (selection conditions). For such selection conditions, correlation analysis can be performed from the perspective of whether there is a significant difference in the evaluation values ​​obtained under different conditions. Specifically, if conditional information is obtained by dividing the time of day into three periods: morning, afternoon, and night, a correlation can be considered to exist if the difference in the average evaluation values ​​for each time period is significant.

[0045] Furthermore, conditional information indicating the subject's mental state includes arousal level and emotional valence. That is, the subject's mental state can be represented two-dimensionally on two axes: arousal level and emotional valence. Arousal level is an indicator representing the subject's level of consciousness and may be measured based on biometric indicators such as heart rate and eye movements, or obtained through the subject's self-reporting. Emotional valence is an indicator representing the subject's emotional state and may be estimated based on biometric indicators such as electroencephalograms, skin potential, skin temperature, and heart rate variability, or obtained through the subject's self-reporting. In this embodiment, a numerical value representing the subject's degree of pleasure or displeasure is used as emotional valence. In this embodiment, arousal level and emotional valence are each represented as integers between 0 and 100, but the minimum and maximum values ​​can be arbitrarily defined.

[0046] [1] Correlation analysis method between evaluation conditions and subjective evaluation The analysis method of this embodiment will be described in detail below with reference to the drawings. Figure 1 is a flowchart of the correlation analysis procedure. First, in steps S11 and S12, condition information and evaluation values ​​are obtained for each evaluation term and object. Note that the order of steps S11 and S12 is not limited. Figure 2 is a diagram showing an example of the information to be obtained. Here, information on age and gender, arousal level and emotional valence indicating mental state are used as condition information indicating the subject's state conditions, and information indicating the time of day is used as condition information indicating environmental conditions. The evaluation value is a value that represents a subjective evaluation indicating the strength of the subject's feeling towards the object, using a specific evaluation term as an indicator, and is obtained by the subject responding based on their own senses after coming into contact with the object. Here, the sense is a sense obtained by the usual way of coming into contact with the object, and is selected from touch, sight, smell, taste, and hearing. In this way, multiple pieces of condition information and evaluation values ​​are obtained in association for each evaluation term and object.

[0047] In this invention, steps S11 and S12 are performed under multiple conditions by changing the conditions. Regarding the acquisition of information under multiple conditions, naturally changing conditions may be utilized, or changes in conditions may be intentionally induced. One way to intentionally induce changes in conditions is, for example, to change the state conditions of arousal level and emotional valence by providing stimuli to the subject, thereby acquiring condition information and evaluation values ​​in steps S11 and S12. As stimuli, for example, images that induce a specific mental state may be provided as visual stimuli. In addition, stimuli such as sounds, objects, tastes, and smells that induce a specific mental state may be provided as auditory, tactile, gustatory, and olfactory stimuli, respectively, to intervene in the subject's state conditions.

[0048] In step S13, a correlation analysis is performed between each condition information and evaluation value for a specific object. Then, in step S14, combinations of evaluation values ​​and condition information that are recognized as having a statistical correlation are identified. Specifically, a correlation analysis between a certain condition information and multiple evaluation values ​​identifies evaluation terms that have a correlation with that condition information (evaluation values ​​that fluctuate in relation to differences in condition information), and a correlation analysis between evaluation values ​​that use a certain evaluation term as an indicator and multiple condition information identifies condition information that has a correlation with that evaluation value.

[0049] This makes it possible to identify evaluation terms that correlate with specific conditional information and evaluation values, such as which evaluation terms correlate with the emotional valence of a subject. Similarly, it makes it possible to identify conditional information that correlates with evaluation values ​​using specific evaluation terms, such as which conditional information correlates with evaluation values ​​using the evaluation term "happy." According to the present invention, it becomes possible to objectively analyze the relationship between various conditions such as the environment and the subject's state, and the subject's subjective evaluation of an object, which could not be clearly shown in the past.

[0050] For example, by analyzing a specific product, it becomes possible to objectively clarify under what conditions and using what evaluation terms the evaluation value of that product changes. Therefore, it becomes possible to promote products using words, such as "a product that feels good at night" or "a product that gives you a sense of quality when you're feeling down."

[0051] For example, if a correlation is found between specific conditional information and evaluation values ​​using specific evaluation terms as indicators for a particular product, it is conceivable that products with different correlations, using the same conditional information and evaluation values, could be proposed as a set product. Specifically, if the results show that cosmetic product A increases the evaluation value of sensitivity using the evaluation term "like" as an indicator in the morning, while the results show that cosmetic product B increases the evaluation value of sensitivity using the evaluation term "like" as an indicator at night, then it would be possible to propose cosmetic products A and B as a set product, to be used separately in the morning and at night.

[0052] [2] A method of analyzing the properties of an object that causes changes in subjective evaluations of the object due to changes in specific conditions. The present invention relates to a method for analyzing objects whose subjective evaluation fluctuates under specific conditions, based on the correlation between evaluation values, which are indexed using evaluation terms, and conditional information. Figure 3 is a flowchart illustrating the analysis method in this embodiment and the procedure for designing an object using this method. Here, as an example, the procedure for analyzing the properties of an object based on evaluation values ​​that fluctuate depending on the time of day, and designing a product using the results, is shown.

[0053] In step S21, an evaluation term whose evaluation value fluctuates under specific conditions is determined as a specific term for a particular object. The detailed procedure for step S21 is the same as the procedure shown in Figure 1. Here, time of day is used as condition information to determine an evaluation term whose evaluation value correlates with the time of day. In other words, an evaluation term is determined as a specific term in which the average evaluation value shows a significant difference in any combination of morning and afternoon, afternoon and night, or night and morning.

[0054] Next, in step S22, evaluation values ​​using the specific term determined in step S21 as an indicator are obtained from multiple subjects for multiple objects of the same type as the object in step S21 but with different properties, while varying the time period (condition information that correlates with the evaluation value using the specific term as an indicator). At this time, it is preferable that the other condition information is not varied or is distributed so as not to be affected by variations in other condition information.

[0055] Here, "of the same type as one object" means that the objects have at least the same use. The scope of what constitutes "of the same type" can be determined arbitrarily. For example, if one object is "cream," then lotions and foundations, which have a common use from the perspective of being cosmetics applied to the skin, may be included as objects of the same type. Alternatively, since cream and lotions or foundations are generally treated as separate products, only "cream" may be included in the analysis of objects of the same type.

[0056] In step S23, multiple objects whose evaluation values ​​changed above a threshold in response to variations in time of day are identified. The threshold may be set, for example, based on the magnitude of the change in the evaluation value using a specific term as an indicator over time, or it may be set based on the standard score or ranking among multiple objects regarding the magnitude of that change. This identifies objects among multiple objects whose evaluation values ​​using a specific term as an indicator fluctuated significantly, either absolutely or relatively, over time.

[0057] In the subsequent step S24, the properties common to the objects identified in step S23 are identified. Here, properties refer to objective information that indicates the characteristics of the object itself, such as mechanical properties such as hardness, elasticity, and viscosity; chemical properties such as volatility, scent, and composition; and optical properties such as color and transparency. Information indicating these properties is recorded as property information for each object, and the analysis is performed based on this information.

[0058] Furthermore, in step S25, the product can be designed to have the properties identified in step S24. For example, if multiple objects in step S24 all have a certain level of elasticity, the target elasticity can be determined from those objects, and the raw materials, composition, and processing can be set to satisfy that target, and the product can be designed accordingly. The target properties can be determined in any way from the properties of multiple objects, based on the average or minimum values ​​of the properties of the multiple objects identified in step S23. Note that multiple common properties can be identified in step S24. In that case, in step S25, the product can be designed to satisfy all the conditions related to the multiple properties.

[0059] Here, we will explain a specific example of designing a lotion whose feel changes depending on the time of day. Let's assume that, given a lotion as a single object, there is a correlation between the time of day and the evaluation value using the evaluation term "like". In this case, in step S22, evaluation values ​​are obtained from multiple subjects for several lotion samples, each with different properties, using the evaluation term "like" as an indicator, while varying the time of day.

[0060] Then, from the acquired data, multiple samples of lotions (of the same type as the target product) that show significant fluctuations in evaluation values ​​based on the evaluation term "liking" depending on the time of day are identified (Step S23), and common properties among these samples are identified (Step S24). The analysis here can be carried out from various perspectives, such as whether there are commonalities in ingredients and fragrance (chemical properties), whether there are commonalities in color and transparency (optical properties), or whether there are commonalities in physical properties such as viscosity and elasticity (mechanical properties).

[0061] For example, if high transparency is identified as a common property, the target transparency can be determined from the transparency of the sample, and the product can be designed to satisfy that property. The sample may also include lotions, creams, and other products with common uses in a broad sense, in addition to toners.

[0062] In this way, by determining specific terms whose evaluation values ​​are expected to fluctuate based on the target condition information, and further identifying the common properties of objects with large fluctuations, a product design method can be provided that realizes fluctuations in evaluation values ​​according to the conditions, from the perspective of product properties.

[0063] Furthermore, a design method for designing an object using the analysis method of the present invention may include a product packaging design step. Specifically, the packaging design step may include attaching associated information to specific terms identified in step S21 and conditions that correlate with evaluation values ​​using those specific terms as indicators.

[0064] This allows for, for example, a correlation between time of day and the evaluation value of the evaluation term "like," where the evaluation value of "like" increases at night. In such cases, packaging can be designed with information such as "You'll like it at night," enabling promotion with substantiated language. It should be noted that in the packaging design process, it is not always necessary to use specific terms directly. Synonyms of specific terms, other words that give a similar impression, or images can be used as information that associates the specific terms and conditions.

[0065] Furthermore, in step S22, evaluation values ​​can be obtained using various objects regardless of their type, and in step S23 and beyond, the relationship between the properties of the objects and the correlation between the evaluation values ​​and condition information can be analyzed. The evaluation terms and condition information used here can be arbitrary. For example, it is conceivable to use various physical properties related to friction, roughness, temperature, deformation, etc., as properties of the objects, and analyze the relationship between those physical properties and the correlation values ​​between the evaluation values ​​and condition information. Examples of correlation values ​​between evaluation values ​​and condition information include the correlation coefficient, p-value, the number or percentage of subjects who showed a correlation, etc.

[0066] In this way, it becomes possible to estimate the properties of an object that show a desired correlation between evaluation values ​​and specific condition information using specific evaluation terms. Conversely, it becomes possible to estimate information regarding the correlation between evaluation values ​​and specific condition information using specific evaluation terms as indicators for that object, based on its properties.

[0067] This allows us to estimate the properties of an object that exhibit the desired correlation, so in step S25, we can design the product with those properties as the target. More specifically, for example, we can estimate the physical properties of an object that show a predetermined correlation coefficient between an evaluation value using the evaluation term "like" as an index and an emotional valence indicating the degree of pleasure or displeasure. By designing the product to have these physical properties, it becomes possible to design a product that produces a predetermined correlation coefficient between the evaluation value using the evaluation term "like" and the emotional valence (condition information). It is also possible to estimate physical properties that do not show a correlation between specific evaluation values ​​and condition information, so that the feeling does not change under specific conditions, and then develop the product with those physical properties as the target.

[0068] [3] Method and apparatus for classifying the variability characteristics of subjective evaluations This invention relates to a classification method for classifying subjects based on the variability characteristics of their subjective evaluations. In this invention, the variability characteristics of specific conditional information and evaluation values ​​that are intercorrelated are analyzed for each subject. For example, various variability characteristics are possible, such as whether the evaluation value increases or decreases as the conditional information increases, and the slope of that increase. This invention assumes that such variability characteristics differ among subjects and provides a method for classifying subjects based on these variability characteristics, and a method for determining recommended products based on that classification.

[0069] In this invention, evaluation values ​​are obtained multiple times from the same subject, using the same object and the same evaluation term, but with different specific condition information that correlates with the evaluation value in that evaluation term. Then, by analyzing the relationship between the changes in the condition information and the changes in the evaluation value, subjects are classified according to their variability characteristics.

[0070] Herein, the present invention is executable by a computer. Figure 4 is a functional block diagram of a classification device 1 that executes the classification method of the present invention. As the classification device 1, a general computer device such as a server can be used, which is equipped with arithmetic units such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), main memory such as RAM (Random Access Memory), auxiliary storage such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), or flash memory, and various input / output devices including means for connecting to a network. By installing a program that allows the computer to function as each means of the classification device 1 in the present invention into a storage device, any computer can function as the classification device 1 of the present invention.

[0071] The classification device 1 comprises a classification means 11, a product suggestion means 12, and a database 13. The classification means 11 acquires one or more evaluation values ​​and condition information of a target and classifies the target based on their variability characteristics. Methods for classifying the target include acquiring multiple evaluation values ​​and condition information for multiple subjects and performing cluster analysis. Alternatively, classification may be performed based on the sign, slope, or other characteristics of the fluctuation in evaluation values ​​in response to fluctuations in condition information. For each target to be classified, the classification is determined by acquiring evaluation values ​​and condition information at the time of evaluation, based on their combination. For example, by acquiring only one evaluation value under specific conditions, the classification to which the combination is expected can be identified based on the distance from the evaluation value under those conditions in each classification. Alternatively, multiple combinations of condition information and evaluation values ​​can be acquired, and the classification of the target can be determined by their fluctuations.

[0072] The product suggestion means 12 suggests recommended products based on the classification of the target person identified by the classification means 11. Specifically, recommended products for each classification are stored in the database 13 in advance, and the product suggestion means 12 can determine recommended products by referring to the database 13 based on the classification. In addition to classifying the variable characteristics, information on the target person's attributes and preferences may also be acquired and recommended products may be suggested taking these into consideration. For example, the system may be configured to ask the target person about their preferences regarding the variation in user experience according to specific conditions, and then suggest products that exhibit the desired variation in evaluation values ​​within the target person's classification.

[0073] In addition to recording the evaluation values ​​shown in Figure 2, database 13 stores recommended products for each category. Furthermore, when considering information on the attributes and preferences of the target audience in addition to the classification of variable characteristics, it is preferable to store the classification of variable characteristics to which the recommendation applies, as well as the attributes and preferences, for each product.

[0074] Figure 5 is a flowchart showing the procedure for the classification method in this embodiment. First, in step S31, the classification means 11 analyzes the condition information and the variation characteristics of the evaluation value for each subject for a specific combination of evaluation terms and conditions. This defines the classification of the variation characteristics. Note that classifications may be defined for multiple combinations of evaluation terms and conditions.

[0075] Next, in step S32, one or more combinations of condition information and evaluation values ​​that serve as the basis for classification are obtained for each subject. In the following step S33, the classification means 11 identifies the classification of each subject based on the variation characteristics of the condition information and evaluation values, based on the information obtained in step S32. Then, in step S34, the product suggestion means 12 determines recommended products by referring to the database 13 based on the classification of the subject, outputs them to the subject, and ends the process.

[0076] As described above, by proposing products according to the classification of target individuals based on their condition information and the characteristics of their evaluation value fluctuations, it is possible to consider the differences in subjective evaluation fluctuations according to the conditions and use this information to propose products that will lead to higher customer satisfaction.

[0077] [4] Method and apparatus for estimating evaluation values ​​and evaluation conditions for a given object using a certain evaluation term as an indicator. The present invention relates to a method and apparatus for estimating condition information from an evaluation value, which is an indicator of an evaluation term, for a given object, based on the correlation between the evaluation value, which is an indicator of an evaluation term, and condition information. In particular, the present invention relates to a method and apparatus for estimating the state of a subject based on an evaluation value, which is an indicator of an evaluation term, for a given object, based on the correlation between the evaluation value, which is an indicator of an evaluation term, and condition information, from condition information. In the present invention, as described above, one of the evaluation value and the condition information is estimated from the other based on the correlation identified by correlation analysis of multiple acquired evaluation values ​​and condition information.

[0078] Herein, the present invention is executable by a computer. Figure 6 is a functional block diagram of the estimation device 2 that executes the estimation method of the present invention. As the estimation device 2, a general computer device such as a server can be used, which is equipped with arithmetic units such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit), main memory such as RAM (Random Access Memory), auxiliary storage such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), or flash memory, and various input / output devices including means for connecting to a network. By installing a program for making the computer function as each means of the estimation device 2 in the present invention into a storage device, any computer can function as the estimation device 2 of the present invention.

[0079] The estimation device 2 comprises an acquisition means 21, an evaluation value estimation means 22, a state estimation means 23, a suggestion means 24, and a database 25.

[0080] The acquisition means 21 acquires either a subjective evaluation value based on specific evaluation terms as an indicator, or conditional information at the time of estimation, from the subject of estimation. The evaluation value or conditional information acquired here has a correlation with the conditional information or evaluation value to be estimated, and this correlation is stored in the database 25.

[0081] The evaluation value estimation means 22 estimates the subjective evaluation value of the subject based on the condition information acquired by the acquisition means 21 and the correlations stored in the database 25.

[0082] The state estimation means 23 estimates information indicating the subject's state, namely the subject's attributes and mental state, based on the evaluation values ​​acquired by the acquisition means 21 and the correlations stored in the database 25.

[0083] For example, if there is a correlation between the evaluation value using the evaluation term "like" for a cream and the subject's emotional valence, and the database 25 stores this correlation, the acquisition means 21 can acquire the emotional valence, and the evaluation value estimation means 22 can estimate the evaluation value using the evaluation term "like" based on the correlation. Also, if the acquisition means 21 acquires the evaluation value using the evaluation term "like," the state estimation means 23 can estimate the subject's emotional valence based on the correlation.

[0084] The suggestion means 24 determines a suggestion for the subject based on the estimation results from the evaluation value estimation means 22 or the state estimation means 23. For example, if the evaluation value estimation means 22 estimates that the subject is likely to feel "like" the cream based on the correlation in the above example, it is assumed that the suggestion means 24 will suggest using the cream. In this way, the suggestion means 24 identifies and outputs actions or products that the subject is likely to find pleasant (likely to have a positive impression of) based on the estimated evaluation value, by referring to the database 25.

[0085] Database 25 stores the correlation between evaluation values ​​and condition information. The correlations include information such as formulas (functions of evaluation values ​​and condition information) that show the correlations obtained through correlation analysis of evaluation values ​​and condition information. It also stores the estimation results from the evaluation value estimation means 22 or the state estimation means 23 in association with the suggestions. This allows the suggestion means 24 to identify the suggestion corresponding to the estimation result.

[0086] Figure 7 is a flowchart illustrating the procedure for the estimation method in this embodiment. Figure 7(a) shows the procedure for estimating the evaluation value, and Figure 7(b) shows the procedure for estimating the subject's condition.

[0087] In estimating the evaluation value, first, in step S41, the acquisition means 21 acquires condition information that has a correlation with the evaluation value to be estimated. Next, in step S42, the evaluation value estimation means 22 estimates the evaluation value from the condition information based on the correlation stored in the database 25. Then, in step S43, the proposal means 24 refers to the database 25 and determines a proposal corresponding to the estimation result.

[0088] Similarly, in the estimation of a state, first in step S51, the acquisition means 21 acquires evaluation values ​​of evaluation terms that have a correlation with the state to be estimated. Next, in step S52, the state estimation means 23 estimates the state from the condition information based on the correlations stored in the database 25. Then, in step S53, the proposal means 24 refers to the database 25 and determines a proposal corresponding to the estimation result.

[0089] Furthermore, the database 25 may also store the relationship between physical property values ​​indicating the properties of an object, and the correlation between evaluation values ​​and condition information, without limiting it to specific objects. In this case, in the steps prior to steps S41 and S51, the acquisition means 21 acquires physical property values ​​indicating the properties of the object to be estimated, and in subsequent processing, the evaluation value or state is estimated based on the correlation between evaluation values ​​and condition information corresponding to those physical property values. This makes it possible to estimate evaluation values ​​or states based on the correlation between evaluation values ​​and condition information, without limiting it to specific objects, based on their properties.

[0090] As described above, by estimating the subject's condition and subjective evaluation, and making suggestions based on those results, it becomes possible to make suggestions that have a more effective positive impact on the subject.

[0091] The present invention will be further explained below with reference to examples of object selection, evaluation terminology selection, acquisition of data such as evaluation values ​​and condition information, and correlation analysis, but it goes without saying that the present invention is not limited to these examples. [Examples]

[0092] [1] Selection of candidate objects In this example, the aim was to examine the sense of touch as one example of human senses. In order to examine the sensations of a wide range of substances that are touched in daily life, we selected the following 34 candidate objects to be evaluated, based on the report by Jeremy A. Fishel et al. (Jeremy A. Fishel et al., frontiers in neurorobotics, published: 18 June 2012), which comprehensively analyzed the physical characteristics of 117 types of substances using a finger-shaped pressure, temperature, and vibration measuring device.

[0093] Lotion, serum, cream, foundation, adhesive tape, slime, clay, human hair, artificial skin, fur, coarse filter, extra-fine filter, sponge rubber sheet, rubber sheet, low-rebound urethane, broadcloth, cotton, nylon, cashmere, leather (synthetic), polystyrene foam, Western paper, Japanese paper, wooden board (cypress), stainless steel sheet, polishing sponge, cork, ribbed rubber, tile, acrylic sheet, artificial leather, lotion (after application), serum (after application), cream (after application).

[0094] [2] Selection of candidate evaluation terms While there have been reports on terminology used for groups of materials within the same category, a systematic glossary of terms for evaluating tactile sensation across a wide range of materials using the same evaluation criteria had not been created. Therefore, we selected candidate evaluation terms as follows.

[0095] (1) Higher-order and lower-order evaluation terms Based on the finding that "the cognitive process of material feel is structured hierarchically" (Kensuke Tobitani, Noriko Nagata, Journal of the Institute of Image Information and Television Engineers, Vol. 71 (2017), No. 11), a wide variety of words expressing sensation, especially touch, were classified into multiple hierarchies. Specifically, evaluation terms that directly describe the physical characteristics of an object, such as "smooth" and "cold," were defined as evaluation terms representing lower-level sensitivities, while evaluation terms that express impressions evoked from the physical characteristics of an object or lower-level sensitivities, such as "comfort" and "luxury," were defined as evaluation terms representing higher-level sensitivities. In this example, a glossary of evaluation terms was created based on this lower-level / higher-level classification.

[0096] (2) Extraction of candidate evaluation terms First, tactile evaluation samples were presented to the subjects, and they were asked to freely describe the feel of the samples and the feelings and emotions they experienced at the time, thereby extracting potential evaluation terms. In this example, the test was conducted on 28 subjects aged 20-59, with an equal proportion of men and women and different age groups. Thirty-four different types of tactile evaluation samples were prepared, and each subject was asked to evaluate 12 samples. The test was conducted with the subjects' eyes open, but before the test, the subjects were instructed not to include any information other than tactile information, and any information related to senses other than touch that was included in their descriptions was deleted.

[0097] From the text obtained as described above, words that were thought to have been obtained through touch were extracted. Furthermore, from the extracted words, terms that were reported in a previous study (Keitaro Kuramitsu, Journal of the Japan Society of Cosmetic Chemists, Vol. 49 (2015), No. 4, pp. 319-327) that describes the physical characteristics of a substance and creates a glossary of terms for evaluating touch using the same evaluation axis were classified as evaluation terms representing lower-order sensitivities. In addition, terms other than evaluation terms representing lower-order sensitivities, and related to impressions and emotional recall, were classified as evaluation terms representing higher-order sensitivities.

[0098] As a result, including duplicates of similar terms, 1,000 evaluation terms representing lower-level sensitivities and 3,000 evaluation terms representing higher-level sensitivities were extracted. From these, 30 candidate terms for lower-level tactile sensitivity and 40 candidate terms for higher-level tactile sensitivity were selected, based on the frequency of occurrence between subjects and between samples, and referencing previously published literature on the feel of materials, a dictionary of emotional expressions (by Akira Nakamura), and a classified vocabulary list (National Institute for Japanese Language and Linguistics).

[0099] [3] Selection of objects and evaluation terms From the candidates for objects and evaluation terms selected in [1] and [2] respectively, the most appropriate objects and evaluation terms were selected as follows.

[0100] (1) Subjects The study included 64 participants aged 20-50, with an equal representation of men and women and different age groups.

[0101] (2) Number of trials Each participant was asked to rate the appropriateness and intensity of 70 candidate evaluator terms for each presented sample using a 6-point scale. Appropriateness refers to whether the participant felt the evaluator term was appropriate as an evaluation axis for the sample, while intensity refers to the strength of the impression or feeling the evaluator term conveyed to the sample. Each participant was presented with a maximum of six samples, and the order of the samples and terms was randomized. Each sample was required to receive at least 10 responses. The test duration was limited to 90 minutes.

[0102] (3) Test conditions Participants were instructed to touch all samples with their eyes open. Furthermore, the touch area was limited to the pad of the index finger of the non-dominant hand, and participants were required to always be touching the sample while evaluating the terms. Participants responded to the terms displayed on the PC monitor using their dominant hand via a mouse. They wiped their fingers between each sample exchange to ensure consistent finger condition at the start. Prior to the test, participants were instructed to practice all operations and to avoid making judgments based on non-tactile factors such as sight or smell.

[0103] (4) Selection of objects and evaluation terms For each sample and term, the mean and standard deviation of the "appropriateness" and "intensity" ratings for each subject were calculated. Furthermore, to explore the similarity of characteristics between samples and evaluation terms and to use this as a reference for selection, cluster analysis was performed using hierarchical clustering methods with the "intensity" and "appropriateness" ratings. The number of clusters was determined by the silhouette coefficient, with terms classified into 6 clusters and samples into 4 clusters based on the number of classifications with the highest silhouette coefficient. Cluster classification was performed using Python and the modules matplotlib.pyplot, scipy.cluster.hierarchy, and sklearn.metrics.

[0104] The target objects were selected based on the following criteria: an overall average appropriateness score of 3.5 or higher for each sample type, and an average intensity score of 2.5 or higher under the same conditions. Selection was also made considering usage examples in previous studies and the balance between clusters in the results of cluster analysis. As a result, nine objects were selected as targets: rubber sheet, stainless steel sheet, low-rebound urethane, cream, coarse filter, polishing sponge, artificial skin, cashmere, and beauty serum.

[0105] The evaluation terms were selected based on the following criteria: an overall average appropriateness score of 3.0 or higher for each term across subjects and samples, and an average intensity score of 2.5 or higher under the same conditions. The terms were chosen after comprehensively considering their usage history in previous studies, the balance between clusters in the cluster analysis results, and the balance between low- and high-order terms to be approximately 1:1 to 1:2. As a result, the following 40 terms were selected as evaluation terms for both the low- and high-order categories.

[0106] Low-level evaluation terms: smooth, silky, slippery, soft, moist, skin-friendly, silky, elastic, warm, supple, hard, rough, dry, rough, bumpy, sticky, cold, prickly Higher-level evaluative terms: pleasant, comfortable, luxurious, elegant, delicate, gentle, light, like, high quality, happy, calming, warm, refreshing, heavy, delighted, fun, surprised, irritated, dislike, sad, scary

[0107] [4] Data acquisition (1) Subjects The study targeted 20 individuals aged 20-50, with an equal representation of men and women and different age groups.

[0108] (2) Test conditions Using the evaluation terms and objects determined in [3], the intensity was assessed using the Visual Analogue Scale (VAS) method in the same test procedure. Before the test, participants practiced answering using the VAS method and were instructed not to make judgments based on factors other than touch, such as sight or smell. The test time was kept within 90 minutes per session. The above method was repeated six times per person, with different dates and times. In addition to obtaining the intensity of the evaluation terms for the objects, environmental conditions during the evaluation, such as the time of day (morning, afternoon, evening) and mental state (arousal level and emotional valence), were also obtained. The dates and times were adjusted so that there were two morning, two afternoon, and two evening sessions out of the six. The six tests were conducted within a predetermined 14-day period.

[0109] (3) Acquired data Using the objects and evaluation terms determined in [3], each subject was asked to rate the intensity of the impression or feeling represented by the evaluation term for each object, using the VAS method as described above, and the evaluation values ​​were obtained. At the same time, the time of day for evaluation (morning, afternoon, evening) was recorded as the environmental conditions of the evaluation, and the subject's age, gender, and mental state (arousal level and emotional valence) were recorded as state conditions indicating the subject's condition.

[0110] [5] Correlation analysis Correlation analysis was conducted on evaluation values ​​for various conditions and evaluation terms, and the following correlations were observed between the conditions and evaluation terms. Furthermore, significant correlations were suggested between evaluation values ​​for multiple other conditions and evaluation terms. (1) Condition: Emotional valence (0-100) Object: Cream Evaluation term: Like Correlation coefficient: 0.3479 (2) Condition: Age (20-59) Object: Rubber sheet Evaluation term: dislike Correlation coefficient: 0.5150 (3) Condition: Gender Object: Low-rebound sponge Evaluation term: sense of luxury p-value: 0.00015 (4) Conditions: Time of day (morning, afternoon, evening) Target: Artificial skin Evaluation term: Happy Significant difference: A significant difference in evaluation values ​​was observed between the morning-afternoon and afternoon-evening periods.

[0111] [6] Conclusion From the above, it became clear that there are combinations of evaluation values ​​(using evaluation terms as indicators) and evaluation conditions in which a correlation can be observed. Furthermore, it was suggested that the relationship between sensation and evaluation conditions can be analyzed using a similar method. [Examples]

[0112] Next, the present invention will be explained in more detail by citing an example that analyzes the relationship between the correlation between evaluation values ​​and evaluation conditions using evaluation terms as indicators, and the nature of the object of evaluation. It goes without saying that the present invention is not limited to this example.

[0113] [1] Measurement of the properties of the object This example aims to examine the sense of touch as one example of human senses. To investigate the sensations associated with a wide range of materials encountered in daily life, a finger-shaped measuring device (Syntouch's Toccare system) capable of measuring pressure, temperature, and vibration was used to measure 15 types of physical properties for five types of objects: cashmere, abrasive sponge, artificial skin, beauty serum, and cream. These physical properties were selected primarily to influence the sense of touch.

[0114] fST: Tactile Stiction, fRS: Sliding Resistance, mTX: Macrotexture, mCO: Macrotexture Coarseness, mRG: Macrotexture Regularity, μRO: Microtexture Roughness, μCO: Microtexture Coarseness, aTK: Adhesive Tack, tCO: Thermal Cooling, tPR: Thermal Persistence, cCM: Tactile Compliance, cDF: Local Deformation, cDP: Damping, cRX: Relaxation, cYD: Yielding.

[0115] [2] Selection of evaluation terms In this embodiment, we further narrowed down the evaluation terms selected in Example 1 to 22 terms. For the lower-order evaluation terms, we selected those that are more directly related to physical properties, and for the higher-order evaluation terms, we selected those that are more likely to be related to emotional valence (expressing pleasure / displeasure) and arousal level. Specifically, the following evaluation terms were selected.

[0116] Ten words—"soft," "hard," "smooth," "rough," "slippery," "bumpy," "cold," "warm," "moist," and "sticky"—were defined as evaluation terms representing lower-level sensitivities that directly describe the physical characteristics of an object. In addition, twelve words—"luxury," "pleasant," "gentle," "warmth," "like," "high quality," "happy," "calming," "comfortable," "irritating," "disgusting," and "dislike"—were defined as evaluation terms representing higher-level sensitivities that describe impressions evoked by the physical characteristics of an object and lower-level sensitivities. In this example, a glossary of evaluation terms was created based on this lower-level / higher-level classification.

[0117] [3] Data acquisition (1) Subjects The study included 46 healthy Japanese individuals aged 20-50, with an equal representation of men and women and different age groups.

[0118] (2) Changes in the subject's condition due to stimulation In this embodiment, the state conditions were changed by applying stimuli to the subjects, and the same subjects were repeatedly asked to provide evaluation values ​​for each object using the same evaluation terms as indicators. Specifically, five types of images were selected from each of the five quadrants—anger, happiness, calmness, relaxation, and sadness—from the OASIS emotion-evoking stimulus image database, and these were presented to the subjects using a display before they touched the objects to be evaluated.

[0119] (3) Acquisition of emotional valence and arousal After the stimulus was presented, participants were asked to use an effective slider to rate their current state, indicating whether their emotional valence (representing pleasure or displeasure) or arousal level (representing calmness or alertness) was closer to their current state. This allowed us to obtain the participants' emotional valence and arousal level as conditional information representing their state, using a 101-point scale from 0 to 100.

[0120] (4) Obtaining evaluation scores Each participant was asked to rate the intensity of 22 evaluative term candidates for one presented sample using the Visual Analogue Scale (VAS). Prior to the test, participants practiced using the VAS and were instructed not to rely on factors other than touch, such as sight or smell. In this context, intensity refers to the degree to which the participant felt the impression or sensation represented by the evaluative term in relation to the sample.

[0121] In this experiment, participants were instructed to place all samples in black boxes, preventing them from seeing the samples, and to touch them under these conditions. Furthermore, they were limited to touching only the pad of the index finger of their non-dominant hand, and were required to always be touching the sample while evaluating the terms. Participants answered the terms by operating a mouse with their dominant hand and responding to the terms displayed on the PC monitor. They wiped their fingers between each sample exchange to ensure the starting condition of their fingers was consistent. Prior to the test, participants were instructed to practice all operations and not to rely on factors other than touch, such as sight or smell, when making judgments.

[0122] [4] Correlation analysis Correlation analysis was performed for each subject between psychological scores, expressed as a combination of emotional valence and arousal level, and evaluation values ​​for each evaluation term. For each subject, the correlation coefficient and p-value were calculated for the correlation between psychological scores and evaluation values ​​for each object.

[0123] Furthermore, for each object, the results of multiple subjects were categorized as either positive or negative based on the correlation coefficient, and the average correlation coefficient for each subject was calculated. In addition, subjects whose absolute value of the correlation coefficient between psychological scores and evaluation values ​​was 0.4 or higher were judged to have a correlation, and the number of subjects with a correlation was identified for each object.

[0124] Furthermore, correlation analysis was performed on each of the correlation coefficients and the number of subjects with correlation, based on the physical properties of the object. The results of the correlation analysis between each of the 15 physical properties listed above for each object, the correlation coefficient, and the number of subjects with correlation showed a significant correlation between the physical properties and the psychological scores and evaluation values ​​in the combinations shown in Figures 8 and 9.

[0125] Figure 8 shows the average correlation coefficients between emotional valences representing pleasure and displeasure and evaluation values ​​for the evaluation term "pleasant," and combinations where a certain level of correlation was observed between these values ​​and each physical property. The average correlation coefficient represents the average of the positive and negative values ​​that showed correlation.

[0126] Figure 9 also shows the number of subjects who showed a correlation between emotional valence (expressing pleasure / displeasure) and evaluation values ​​for various evaluation terms, and the combinations of each physical property that showed a correlation of a certain level or higher.

[0127] [5] Conclusion From the above, it became clear that there are combinations of information regarding the correlation between evaluation values ​​and evaluation conditions (particularly the mental state of the subject) using evaluation terms as indicators, and the physical properties of the object being evaluated, in which a correlation can be observed. Furthermore, it was suggested that the relationship between the correlation between evaluation values ​​and evaluation conditions and the properties of the object being evaluated can be analyzed using a similar method. [Explanation of symbols]

[0128] 1:Classification device 2: Estimation device 11: Classification means 12:Product proposal means 13: Database 21: Acquisition means 22: Evaluation Value Estimation Method 23: State estimation means 24: Proposal means 25: Database

Claims

1. A method for identifying subjective evaluations that vary depending on conditions, performed by a computer, The steps include obtaining subjective evaluation values ​​from subjects, using evaluation terms as indicators, for the object under multiple different conditions, and Steps include obtaining conditional information that indicates the conditions during evaluation, An analysis method comprising the steps of: performing a correlation analysis between specific condition information and the evaluation value for each of several evaluation terms, and identifying evaluation terms that have a correlation with the condition information.

2. The analysis method according to claim 1, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which the subject feels the sensation represented by the evaluation term.

3. The analytical method according to claim 1 or claim 2, wherein the conditions include environmental conditions that represent the environment surrounding the subject.

4. The analytical method according to claim 3, wherein the aforementioned environmental conditions include time periods within a day.

5. The analytical method according to any one of claims 1 to 4, wherein the conditions include state conditions indicating the state of the subject.

6. The analytical method according to claim 5, wherein the aforementioned condition includes the age or age group of the subject.

7. The analysis method according to claim 5 or 6, wherein the aforementioned state conditions include the mental state of the subject.

8. The analysis method according to claim 7, wherein the condition information indicating the state conditions includes emotional valence indicating the emotions of the subject.

9. The analysis method according to claim 7 or claim 8, wherein the condition information indicating the state conditions includes the level of alertness of the subject.

10. A method for identifying conditions that affect subjective evaluation, performed by a computer, The steps include obtaining subjective evaluation values ​​from subjects, using evaluation terms as indicators, for the object under multiple different conditions, and A step to acquire multiple types of conditional information that indicates the conditions during evaluation, An analysis method comprising the steps of: performing a correlation analysis between the evaluation value and condition information in a specific evaluation term for each of several types of condition information, and identifying the condition information that has a correlation with the evaluation value.

11. The analysis method according to claim 10, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which the subject feels the sensation represented by the evaluation term.

12. The analytical method according to claim 10 or claim 11, wherein the conditions include environmental conditions that represent the environment surrounding the subject.

13. The analytical method according to claim 12, wherein the aforementioned environmental conditions include time periods within a day.

14. The analytical method according to any one of claims 10 to 13, wherein the conditions include state conditions indicating the state of the subject.

15. The analytical method according to claim 14, wherein the aforementioned condition includes the age or age group of the subject.

16. The analysis method according to claim 14 or claim 15, wherein the aforementioned state conditions include the mental state of the subject.

17. The analysis method according to claim 16, wherein the condition information indicating the state conditions includes an emotional valence indicating the emotions of a subject.

18. The analysis method according to claim 14 or claim 15, wherein the condition information indicating the state conditions includes the level of alertness of the subject.

19. An analytical method for analyzing the properties of an object, which is performed by a computer and causes changes in subjective evaluations of the object due to changes in specific conditions, The steps include: determining a specific term for evaluation where the evaluation value of a particular object changes depending on specific conditions; The step includes identifying the properties of the object that affect the evaluation value of the aforementioned specific term, The step of determining the aforementioned specific term is: A step of obtaining subjective evaluation values ​​from subjects, using evaluation terms as indicators, for a single object, under multiple different conditions. Steps include obtaining conditional information that indicates the conditions during evaluation, The step includes performing a correlation analysis between specific condition information and the evaluation value for each of several evaluation terms, and determining the evaluation terms for which the evaluation value has a correlation with the condition information as specific terms, The step of identifying the properties of the object is, A step of varying the condition information for multiple objects with different properties and obtaining evaluation values ​​using the specific term as an indicator for each object and condition information, respectively. An analysis method comprising the step of identifying properties that affect the evaluation value of a specific term, based on the properties of multiple objects and the changes in the evaluation value according to the condition information.

20. The step of identifying the properties of the object is, The step of determining the specific term involves varying the condition information for multiple objects of the same type as one object but with different properties, and obtaining evaluation values ​​using the specific term as an indicator for each object and condition information, respectively. The steps include identifying multiple objects from among several objects whose evaluation values ​​have changed to a threshold or higher in response to the change in the condition information, The analytical method according to claim 19, comprising the step of identifying a property common to a plurality of identified objects as a property that affects the specified term.

21. The step of identifying properties that affect the evaluation value of the aforementioned specific term is: A step of analyzing the relationship between the properties of multiple objects with different properties and relational information showing the relationship between the conditional information and evaluation value for each object, The analysis method according to claim 19, comprising the step of identifying properties that affect the evaluation value of the specific term based on the aforementioned relationship.

22. The analysis method according to claim 19, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which the subject feels the sensation represented by the evaluation term.

23. The analytical method according to claim 19 or claim 22, wherein the aforementioned properties include values ​​representing the mechanical properties of the object.

24. The analytical method according to any one of claims 19 to 23, wherein the aforementioned properties include values ​​representing the chemical properties of the object.

25. A classification method performed by a computer for classifying the variability characteristics of subjective evaluations by subjects, A step of obtaining subjective evaluation values ​​from subjects, using evaluation terms as indicators, for an object under specific conditions, Steps include obtaining conditional information that indicates the conditions during evaluation, A classification method comprising the step of identifying a classification of a subject based on the variation characteristics of the evaluation value according to the condition information, based on the correlation between the evaluation value and the condition information, using a combination of the evaluation value and the condition information.

26. The classification method according to claim 25, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which the subject feels the sensation represented by the evaluation term.

27. The steps include storing recommended products based on the aforementioned classification in a database, The classification method according to claim 25 or claim 26, further comprising the step of referring to the database and determining the recommended product based on the classification of subjects according to the variability characteristics.

28. A classification device for classifying the variability characteristics of subjective evaluations by subjects, A means for obtaining subjective evaluation values ​​of an object, using evaluation terms as indicators, under specific conditions, A means of obtaining conditional information that indicates the conditions at the time of evaluation, A classification device comprising means for identifying the classification of a subject based on the fluctuation characteristics of the evaluation value according to the condition information, based on the correlation between the evaluation value and the condition information, using a combination of the evaluation value and the condition information.

29. The classification device according to claim 28, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which a subject feels the sensation represented by the evaluation term.

30. A database that stores recommended products based on the aforementioned classification, The classification apparatus according to claim 28 or claim 29, further comprising means for referring to the database and determining the recommended product based on the classification of subjects according to the variability characteristics.

31. Based on the correlation between the subjective evaluation value of the subject using evaluation terms for the object as indicators and condition information indicating the conditions at the time of evaluation, A computer-based estimation method that estimates subjective evaluation values ​​of an object, using specific evaluation terms as indicators, based on conditional information that indicates the conditions during the evaluation.

32. The estimation method according to claim 31, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which a subject feels the sensation represented by the evaluation term.

33. The aforementioned conditions include state conditions that indicate the subject's condition, The estimation method according to claim 31, wherein the state condition includes, as the mental state of the subject, an emotional valence indicating the subject's emotions.

34. The aforementioned conditions include state conditions that indicate the subject's condition, The estimation method according to any one of claims 31 to 33, wherein the state conditions include the subject's mental state, specifically the subject's level of arousal.

35. The estimation method according to any one of claims 31 to 34, further determining a proposal for the subject based on the estimation result of the aforementioned evaluation value.

36. The estimation method according to any one of claims 31 to 35, wherein the evaluation value is estimated based on property information indicating the properties of the object, in addition to the condition information.

37. Based on the correlation between the subjective evaluation value of the subject using evaluation terms for the object as indicators and condition information indicating the conditions at the time of evaluation, An estimation device that estimates a subject's subjective evaluation value of an object, using specific evaluation terms as indicators, based on condition information that indicates the conditions during the evaluation.

38. The estimation device according to claim 37, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which a subject feels the sensation represented by the evaluation term.

39. The estimation device according to claim 37 or claim 38, which estimates the evaluation value based on property information indicating the properties of the object in addition to the condition information.

40. Based on the correlation between subjective evaluation values ​​of subjects using evaluation terms for the object as indicators and condition information indicating the conditions at the time of evaluation, An estimation method performed by a computer that estimates the state of a subject at the time of evaluation based on subjective evaluation values ​​of the subject, using evaluation terms as indicators for the object.

41. The estimation method according to claim 40, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which a subject feels the sensation represented by the evaluation term.

42. The estimation method according to claim 40 or claim 41, wherein the subject's state includes an emotional valence indicating the subject's emotions as a value indicating the subject's mental state.

43. The estimation method according to any one of claims 40 to 42, wherein the subject's state includes the subject's level of alertness as a value indicating the subject's mental state.

44. The estimation method according to any one of claims 40 to 43, further determining a proposal for the subject based on the estimation result of the subject's condition.

45. Based on the correlation between the subjective evaluation value of the subject, which is an indicator of the evaluation of the object, and the condition information that indicates the conditions at the time of evaluation, An estimation device that estimates the state of a subject at the time of evaluation based on subjective evaluation values ​​of the subject, using evaluation terms as indicators for the object.

46. The estimation device according to claim 45, wherein the evaluation value is a sensitivity evaluation value indicating the strength with which a subject feels the sensation represented by the evaluation term.

47. An estimation device according to claim 45 or claim 46, which estimates the state of the subject based on property information indicating the properties of the object, in addition to the evaluation value.

Images