Evaluation method and evaluation system for products with flavor.

Abstract

This invention provides an evaluation method and system that can visualize the relationship between the aroma and flavor of a product and the emotions evoked by specific aromas and flavors, and to what extent. [Solution] A product evaluation method and evaluation system are provided, which evaluate the relationships between different evaluation images of the flavor of a product, comprising: a first evaluation step of analyzing a first evaluation result that expresses the degree of a first evaluation image evoked by the flavor of the product using one or more numerical items; a second evaluation step of analyzing a second evaluation result that expresses the degree of a second evaluation image evoked by the product using one or more numerical items; and a relationship calculation step of integrating the analysis results of the first evaluation result and the analysis results of the second evaluation result to calculate the relationship between the first evaluation image and the second evaluation image.

Description

Technical Field

[0001] The present invention relates to a method and a system for evaluating products having a fragrance.

Background Art

[0002] Originally, luxury goods such as food and beverages, and daily necessities such as detergents and cosmetics often have their own unique fragrances (scents, aromas, tastes, textures, stimuli, etc.).

[0003] Regarding these fragrances, they may originally be derived from the components contained in the products, but in the luxury goods and daily necessities fields, it is also frequently done to intentionally impart a desired fragrance by adding specific fragrances, seasonings, etc. at the manufacturing stage.

[0004] In recent years, various findings have been accumulated regarding fragrance research, and not only adding specific fragrances and seasonings, but also research on what kind of processing should be performed at the manufacturing stage of products to impart the desired fragrance has been advanced in each field.

[0005] Therefore, in the planning and development stage of products in recent years, not only the main functions of the products, but also the fragrances of the products have become one of the important items regarded as having secondary effects.

[0006] However, since it is difficult to quantify a fragrance, if a fragrance slightly different from the desired fragrance is imparted, or if the fragrance is too strong or too weak, it may conversely worsen the impression of the product. Therefore, various methods for quantifying fragrances and methods for evaluating them in association with other information have been developed.

[0007] As prior art documents for methods of evaluating flavor, for example, Patent Document 1 discloses a method for representing an image in color by creating a color representation diagram corresponding to an image to be visualized through a first, second, and third stage. In the first stage, n subjects select s types of color swatches corresponding to an image from a group of color swatches containing at least a × b swatches, and determine the contribution rate of each selected color swatch. In the second stage, a representation diagram is created for each subject using each color swatch constituting the group of color swatches and the contribution rate determined in the first stage. In the third stage, a total value is calculated by summing the contribution rates included in each subject's representation diagram for each color swatch, and a color representation diagram is created that represents the total value using elements colored to correspond to each selected color swatch, wherein a≧2, b≧2, s≧2, and n≧1 are satisfied in this method for representing an image in color.

[0008] Furthermore, Patent Document 2 discloses a method for extracting scent-related information, characterized by comprising an extraction step in which a model is trained on the probabilistic dependency relationships between (A) multiple numerical items relating to a subject, (B) multiple numerical items relating to the impression of the scent or the mood or psychological state of the subject when smelling the scent, and (C) multiple numerical items relating to scent elements, and the dependency relationships are changed by specifying conditions for some of the numerical items, and extraction items different from the specified items are extracted based on the changed probabilistic dependency relationships. [Prior art documents] [Patent Documents]

[0009] [Patent Document 1] Patent No. 6826195 [Patent Document 2] Japanese Patent Publication No. 2023-20181 [Non-patent literature]

[0010] [Non-Patent Document 1] David A. Aaker, “Managing brand equity: capitalizing on the value of a brand name” (US), Jossey-Bass Inc., 1991 [Overview of the Initiative] [Problems that the invention aims to solve]

[0011] Incidentally, the evaluation (sensory evaluation) of the flavor and aroma of a product when consumed (eating in the case of luxury goods, or using in the case of daily necessities) is frequently conducted during the product development stage. On the other hand, with so many products and services available today, it is difficult to differentiate products solely on "functional benefits such as scent and flavor," as described in Non-Patent Document 1. As a result, there is a growing desire in product development to appeal to consumers and users with "emotional benefits such as scenes and emotions."

[0012] In such cases, if it were possible to connect the flavor (functional value) of a product with the emotions (emotional value; for example, the time of day, place, or mood in which the product would be used), it would be possible to reach the target consumer group more deeply. However, since there are no methods or systems for such evaluation, it often relies mainly on the experience of the product development staff.

[0013] Furthermore, existing methods for evaluating flavor and emotional impact, such as sensory evaluation conducted during the development phase and monitor surveys conducted with consumers after sales, are frequently used because they can quantify flavor and visualize the images evoked by the product.

[0014] In sensory evaluation, since the evaluation is conducted by a specialized panel, the evaluation is rigorous and detailed, making it possible to analyze and express even slight differences between samples in detail. On the other hand, monitor surveys can reflect the genuine voices and evaluations of general consumers. However, until now, there has been no method for analyzing and expressing the relationship between the evaluation results of these two types of evaluations in a way that can be applied to product development.

[0015] As described above, there was a need for specific methods and systems to connect different evaluations, such as the aroma and flavor of a product and the emotions it evokes (finding connections, correlations, and similarities).

[0016] This invention has been made in view of these circumstances, and aims to provide an evaluation method and evaluation system that can visualize the relationship between what emotions are evoked by a particular flavor when the flavor of a product is ingested, and to what extent what flavors are evoked by a particular emotion. [Means for solving the problem]

[0017] The inventors conducted diligent research and developed a product evaluation method for evaluating the relationships between different evaluation images related to the flavors of a product. This method includes a first evaluation step of analyzing a first evaluation result that expresses the degree of the first evaluation image evoked by the product using one or more numerical items, a second evaluation step of analyzing a second evaluation result that expresses the degree of the second evaluation image evoked by the product using one or more numerical items, and a relationship calculation step of integrating the analysis results of the first and second evaluation results to calculate the relationship between the first and second evaluation images. This method makes it possible to visualize the relationship between what emotions are evoked by a particular flavor when the flavors of a product are consumed, and to what extent what flavors are evoked by a particular emotion, thus completing the present invention. In other words, the present invention provides the following:

[0018] (1) The first aspect of the present invention is a method for evaluating a product, which evaluates the relevance between different evaluation images in the fragrance of the product. The method includes a first evaluation step of analyzing a first evaluation result representing the degree of a first evaluation image recalled for the product in one or more quantified items, a second evaluation step of analyzing a second evaluation result representing the degree of a second evaluation image recalled for the product in one or more quantified items, and a relevance calculation step of integratively analyzing the analysis result of the first evaluation result and the analysis result of the second evaluation result to calculate the relevance between the first evaluation image and the second evaluation image.

[0019] (2) The second aspect of the present invention is the evaluation method according to (1), characterized in that the first subjects who evaluate the first evaluation result include professional panelists.

[0020] (3) The third aspect of the present invention is the evaluation method according to (1), characterized in that the first evaluation image includes a fragrance image related to the fragrance recalled by the first subject.

[0021] (4) The fourth aspect of the present invention is the evaluation method according to (1), characterized in that the second subjects who evaluate the second evaluation result include general consumers.

[0022] (5) The fifth aspect of the present invention is the evaluation method according to (1), characterized in that the second evaluation image includes an emotion image related to the emotion recalled by the second subject.

[0023] (6) The sixth aspect of the present invention is the evaluation method described in (1), wherein in the first evaluation step and the second evaluation step, the analysis method of the first evaluation result and the second evaluation result is principal component analysis or correspondence analysis. When the analysis method is principal component analysis, in the correlation calculation step, the correlation between the first evaluation image and the second evaluation image is calculated by a correlation matrix. When the analysis method is correspondence analysis, in the correlation calculation step, the correlation between the first evaluation image and the second evaluation image is calculated by a chi-square distance.

[0024] (7) The seventh aspect of the present invention is an evaluation method for a product that evaluates the relationship between the flavor and emotion in the flavor of the product, comprising a step of analyzing a flavor evaluation representing the degree of a flavor image recalled for the product and an emotion evaluation representing the degree of an emotion image recalled for the product, and calculating the correlation between the flavor evaluation and the emotion evaluation.

[0025] (8) The eighth aspect of the present invention is a product evaluation system that evaluates the correlation between different evaluation images in the flavor of a product, comprising an analysis unit that analyzes a first evaluation result representing the degree of a first evaluation image recalled for the product by one or more quantified items and a second evaluation result representing the degree of a second evaluation image recalled for the product by one or more quantified items, respectively, and an arithmetic unit that integrally analyzes the analysis result of the first evaluation result and the analysis result of the second evaluation result and calculates the correlation between the first evaluation image and the second evaluation image.

Advantages of the Invention

[0026] According to the present invention, it is possible to provide an evaluation method and an evaluation system capable of visualizing the correlation of what emotions are recalled from a specific flavor or what flavors are recalled from a specific emotion and to what extent when ingesting the flavor possessed by a product.

Brief Description of the Drawings

[0027] [Figure 1] This flowchart shows the processing procedure for the evaluation method of a product having flavor according to this embodiment. [Figure 2] This graph plots the results of flavor evaluations conducted by expert panelists on four different coffee samples, using the first and second principal component spaces of principal component analysis. [Figure 3] This graph plots the results of emotional evaluations conducted by general consumers for four different coffee samples, using the first and second principal component spaces of a principal component analysis. [Figure 4] Figures 2 and 3 are graphs plotting the results shown in Figures 2 and 3 onto the first and second principal component spaces of the principal component analysis using the correlation matrix. [Figure 5] This is a two-dimensional map showing the results of a correspondence analysis. [Figure 6] This diagram shows a simplified configuration of an evaluation system for products with flavor. [Figure 7] This graph plots the evaluation results for aroma and emotional (color) of four different shampoo samples in the first and second principal component spaces of a principal component analysis using a correlation matrix. [Figure 8] This graph plots the flavor and emotional evaluation results for four different ice cream samples in the first and second principal component spaces of a principal component analysis using a correlation matrix. [Figure 9] This graph plots the flavor and emotional evaluation results of five different samples of fresh cream in the first and second principal component spaces of a principal component analysis using a correlation matrix. [Figure 10] This graph plots the evaluation results for flavor and emotion of five different air freshener samples in the first and second principal component spaces of a principal component analysis using a correlation matrix. [Figure 11] This graph plots the evaluation results for aroma and emotion of four different shampoo samples in the first and second principal component spaces of a principal component analysis using a correlation matrix. [Modes for carrying out the invention]

[0028] The following describes in detail embodiments for carrying out the present invention (hereinafter simply referred to as "this embodiment"). This embodiment is illustrative for explaining the present invention and is not intended to limit the present invention to the following content. The present invention can be implemented by modifying it as appropriate within the scope of its gist.

[0029] <Evaluation Method for Products with Flavor> Figure 1 is a flowchart showing the processing procedure for the evaluation method of a product having flavor according to this embodiment. The evaluation method according to this embodiment is a method for evaluating products with flavor, which evaluates the relationships between different evaluation images of the flavor possessed by the product.

[0030] In this specification, "flavor" refers to one or more sensations that can be altered by aroma, typically including smell and / or taste. Furthermore, the term "ingesting flavor" includes perceiving sensations including smell and / or taste, as well as sensations other than smell and / or taste, such as coldness, warmth, texture (throat feel, firmness, viscosity, etc., also called texture), and stimulating sensations such as carbonation and spiciness, as a result of ingesting the product. In other words, if the product is food or beverage, flavor includes smell, aroma, taste, texture, and irritation, but even if the product is a household item such as detergent, soap, or cosmetics, flavor includes smell, aroma, texture, and irritation. Furthermore, in this invention, "ingestion" includes not only drinking or eating the product, but also using the product (for example, washing an object if the product is detergent or soap, or applying it to the body if it is a cosmetic).

[0031] The following provides a detailed explanation of each step in the evaluation method.

[0032] (Process S1) Process S1 is the process of selecting products that have flavor. In this specification, "products with flavor" includes, as described above, food and beverages, household goods, etc. Food and beverages include beverages, snacks, sweets, and other products to be drunk or eaten. Household goods include products for use on the body, such as shampoos, soaps, antiperspirant sprays, and body lotions, as well as clothing and household products, such as fabric softeners, laundry detergents, and air fresheners. Furthermore, products include not only commercially available products that are actually on the market, but also prototypes. In the following description, the product having flavor is referred to as "general black coffee (equivalent to commercially available products)," but this is merely an example for illustrative purposes, as described above, and is not intended to limit the present invention to the following.

[0033] (Process S2) Process S2 is the process of selecting the first evaluation image. In this case, the "first evaluation image" is not particularly limited to any visual or auditory expression such as colors, images, words, or sounds that the first subject recalls when consuming black coffee, but it is preferable that it includes a flavor image related to the aroma that the first subject recalls. Since the aroma image is the primary evaluation image, the relationship between aroma and emotion can be visualized more clearly in the evaluation method of the present invention.

[0034] (Process S3) Process S3 is the step of obtaining a response (first evaluation result) that represents the degree of the first evaluation image that comes to mind for the product. In this case, the degree to which the first evaluation image is recalled is expressed by one or more numerical items. These one or more numerical items are, for example, one or more items that represent the language included in the first evaluation image, and each can be evaluated numerically. Furthermore, if the first evaluation image is language, the language used in the evaluation is also called "flavor language (terms that express the flavor characteristics of the product)." Note that the one or more languages ​​may be the same or different for each product with the selected flavor.

[0035] Furthermore, it is preferable that the first subject who answers questions about the degree to which they recall their primary evaluation image includes a well-trained professional panelist (e.g., a flavorist, perfumer, etc.: with 10 years or more of experience). Including a professional panelist allows for a clear evaluation of primary evaluation images, especially those that are difficult to quantify, such as aroma and flavor images. Furthermore, from the viewpoint of ensuring the significance of the evaluation and suppressing excessive variability in the data, it is preferable that the number of first subjects be around 5 to 20.

[0036] (Processing S4) Process S4 is the process of analyzing the first evaluation results. In this context, "analysis" means analyzing the relationship between each quantified item and the average calculated for the entire first group of subjects. At this time, the analytical method is not particularly limited as long as it is a method commonly used in statistical analysis, but multivariate analysis is preferred because it involves multiple items. Examples of analytical methods included in multivariate analysis are principal component analysis, multiple regression analysis, discriminant analysis, correspondence analysis, cluster analysis, multidimensional scaling, covariance structure analysis, and factor analysis, and among these, principal component analysis or correspondence analysis is preferred. By using principal component analysis or correspondence analysis, it is possible to easily find the evaluation item with a high contribution rate among multiple evaluation items in each sample. Note that it is also possible to analyze the first evaluation result before performing the integrated analysis in process S8 described later, but in that case, this process is omitted.

[0037] The above processes S2 to S4 are also referred to as the first evaluation process. To give a specific example of the first evaluation process, for example, the 16 flavor words shown in Table 1 are defined as flavor images, and 10 first-stage subjects (all expert panelists) evaluate the flavors they perceive when drinking room-temperature black coffee, using a line scale for each item. In this case, the weakest is converted to a number of 0, and the strongest to a number of 10. [Table 1]

[0038] The evaluation method in this case is not particularly limited, as long as it is a method normally used by expert panelists, but descriptive analysis is preferred, for example. Furthermore, it is preferable to present the samples by randomizing them with three-digit random codes.

[0039] Next, we analyze the initial evaluation results, in which 10 first-stage subjects evaluated the aroma and taste of each coffee sample. The graph shown in Figure 2 plots the results of expert panelists' evaluations of the aroma and flavor of four coffee samples A-D (referred to as A-D in the figure; the same applies hereafter) on the first and second principal component spaces of principal component analysis. Note that the diagrams in Figure 2 and other graphs are illustrative and not absolute.

[0040] Figure 2 shows the plots for each of the samples A to D, along with the plots for the 16 flavor words and the regions where specific flavor words are concentrated. Among these, sample A shows strong flavors related to grains and nuts, such as "smoky," "bitter," "roasted," "grainy," and "full-bodied / rich."

[0041] Similarly, Sample B has strong aromas related to roast and acidity, such as "fruity," "acidity," "citrus," "smoky," "bitterness," "roast," and "body / richness." Sample C has strong aromas related to citrus and berries, such as "citrus," "crisp / clear," "floral," and "berry." Sample D has strong aromas related to sweetness and chocolate, such as "sweet aroma," "caramel," "sweetness," and "chocolate."

[0042] The above explanation of the first evaluation process uses principal component analysis as an example of analyzing the first evaluation results, but it is not limited to this; for example, correspondence analysis can also be used. In that case, a cross-tabulation table is created by summing the evaluators' evaluation results for each sample and for each flavor image, and this data is then analyzed using correspondence analysis to create a two-dimensional map.

[0043] Furthermore, the results of the first evaluation process described above are also referred to as "flavor evaluation," which represents the degree to which the flavor image evoked by the product is perceived.

[0044] (Process S5) Next, process S5 is the step of selecting the second evaluation image. In this context, the "second evaluation image" is not particularly limited to any visual or auditory expression such as colors, images, words, or sounds that the second subject recalls when consuming black coffee, but it is preferable that it includes emotional images related to the emotions the second subject recalls. In this context, "emotions" include not only the feelings recalled when consuming the flavor of the product, but also visually recalled things such as the scene or image in which the product would be consumed. By making the emotional image the secondary evaluation image, the relationship between flavor and emotion can be visualized more clearly in the evaluation method of the present invention.

[0045] (Process S6) Process S6 is a step in which responses (second evaluation results) that represent the degree of the second evaluation image recalled for the product are obtained. In this case, the degree to which the second evaluation image is recalled is expressed by one or more numerical items. These one or more numerical items are items that represent one or more words included in the second evaluation image, and each can be evaluated numerically. Furthermore, if the second evaluation image is a word, the word used in the evaluation is also called an "emotional word" (a term that represents the feelings recalled when the flavor of the product is consumed, or the visually recalled scenes or images in which the product would be consumed). Note that the one or more words may be the same or different for each product with the selected flavor. Furthermore, as will be discussed later, since the evaluation is conducted by a second group of subjects, including general consumers, the quantifiable items may be those extracted from a pre-survey questionnaire (including the product concept, etc.). Regarding the quantification of the evaluation results, numerical scales such as scoring methods may be used, or frequency data such as the CATA method may be used.

[0046] Furthermore, it is preferable that the second group of participants who respond regarding the degree to which they recall the second evaluation image include general consumers (general panelists). Including general consumers allows for the acquisition of evaluation results that reflect a more popular and realistic perspective. Furthermore, from the viewpoint of ensuring the significance of the evaluation and suppressing excessive variability in the data, it is preferable that the number of second subjects be 50 or more.

[0047] (Process S7) Process S7 is the process of analyzing the second evaluation results. In this case, "analysis" means the same thing as the analysis in process S4. At this time, as with process S4, multivariate analysis is preferred as the method of analysis because it involves multiple items. Examples of analytical methods included in multivariate analysis are principal component analysis, multiple regression analysis, discriminant analysis, correspondence analysis, cluster analysis, multidimensional scaling, covariance structure analysis, and factor analysis, with principal component analysis or correspondence analysis being preferred among them. By using principal component analysis or correspondence analysis, it is possible to easily find the evaluation items with a high contribution rate among the multiple evaluation items in each sample. Note that, as with process S4, the second evaluation result can also be analyzed in process S8, which will be described later, but in that case, this process will be omitted.

[0048] The processes S5 to S7 described above are also referred to as the second evaluation process. In Figure 1, the first evaluation process (processes S2 to S4) and the second evaluation process (processes S5 to S7) are shown as parallel processes, but the order of processing is not particularly limited. Either the first or second evaluation process may be performed first, or they may be performed simultaneously. The following describes a specific example of the second evaluation process. For example, a web-based questionnaire is administered to respondents who drink black coffee, asking "When (place, time of day, environment, mood, etc.) do you drink it?" From the results (441 comments), 10 emotional words shown in Table 2 are extracted and compiled as emotional images. Next, 50 second-party subjects (all general consumers) evaluate the emotions they feel when drinking room-temperature coffee, using a graph scale from 0 to 10 for each item. In this case, the weakest emotion is converted to 0 and the strongest to 10. [Table 2]

[0049] In this case, it is preferable to present the sample by generating a 3-digit random code and then presenting it randomly.

[0050] Next, we analyze the second set of results, in which 50 second-party subjects evaluated the emotional impact of each coffee sample. The graph shown in Figure 3 plots the results of general consumers' emotional evaluations of four types of coffee samples A-D (coffee A-D) in the first and second principal component spaces of a principal component analysis.

[0051] Figure 3 shows plots for each of the samples A to D, along with plots for terms showing significant differences or trends among the 10 emotional words, and regions where specific emotional words are concentrated. Among these, sample A shows a strong desire to drink "when thirsty" or "with a meal."

[0052] Similarly, Sample B strongly evokes feelings of wanting to drink it when one wants to feel a sense of luxury or sophistication, with responses such as "to change one's mood," "with sweets," "when one wants to concentrate," "a sense of luxury," and "a sense of sophistication." Sample C strongly evokes feelings of wanting to drink it to "relieve fatigue," and Sample D strongly evokes feelings of wanting to drink it to "relax," as a "reward," or to "relieve fatigue."

[0053] The above explanation of the second evaluation process used principal component analysis as an example for analyzing the second evaluation results, but it is not limited to this method. For example, correspondence analysis can also be used. In that case, a cross-tabulation table is created by summing the evaluators' evaluation results for each sample and for each emotional image, and this data is then subjected to correspondence analysis to create a two-dimensional map.

[0054] Furthermore, the results of the second evaluation process described above are also referred to as "emotional evaluation," which represents the degree to which emotional images are evoked in relation to the product.

[0055] (Process S8) Finally, process S8 is a step in which the two results, the first evaluation result and the second evaluation result, are integrated and analyzed to calculate the relationship between the first evaluation image and the second evaluation image. Regarding the first and second evaluation results, the obtained results may not be analyzed in processes S4 or S7, but may be analyzed separately before process S8, or all results may be integrated and analyzed in this process, or only a portion of the results may be extracted and analyzed. Furthermore, in this specification, "relationship" may also be referred to as "correlation" or "similarity" depending on the analysis method.

[0056] While the calculation method in this case is not particularly limited, when principal component analysis is used as the analysis method in processes S4 and S7, it is preferable to calculate the relationship (correlation) between the first evaluation image and the second evaluation image using a correlation matrix. Specific examples will be described later, but by calculating using a correlation matrix, the relationship (correlation) between the first evaluation image and the second evaluation image can be visualized numerically. Furthermore, when correspondence analysis is used as the analysis method in processes S4 and S7, it is preferable to calculate the relationship (similarity) between the first evaluation image and the second evaluation image using the chi-squared distance. By calculating the chi-squared distance, the relationship (similarity) between the first evaluation image and the second evaluation image can be visualized numerically.

[0057] First, we will explain using the example of calculating the relationship (correlation) between the first and second evaluation images using a correlation matrix. Figure 4 is a graph plotting the results shown in Figures 2 and 3 in the first and second principal component spaces of the principal component analysis using the correlation matrix. Note that in Figure 4, preference is excluded, and only terms indicating significant difference or trend are listed. As is clear from the results shown in Figure 4, there are correlations between flavors such as "grains" and "nuts" and the emotion of "being thirsty," flavors such as "roasted" and "smoky" and the emotion of "luxury," flavors such as "citrus" and "berries" and the emotion of "relieving fatigue," and flavors such as "sweet scents" and "chocolate" and the emotion of "comfort."

[0058] Furthermore, Table 3 shows the correlation coefficients for each variable, with one of the two variables fixed. In the correlation coefficients presented, a value greater than 0.7 indicates a strong correlation, and a value greater than 0.4 indicates a correlation. [Table 3]

[0059] As shown in Table 3, the sentiment of "with a meal" has a strong positive correlation with flavors such as "nuts" and "grains," and at the same time, a strong negative correlation with the flavor "sour." Furthermore, the sentiment of "to enjoy with sweets" has a positive correlation with aromas such as "sweet scent" and "roast," while simultaneously having a negative correlation with aromas such as "floral," "crisp / clear," "fruity," "berry," "citrus," and "acidity." In this way, by expressing it using a correlation coefficient, the relationship (correlation) between the first evaluation image and the second evaluation image is visualized numerically. In other words, the relationship between the aroma and emotion associated with the product's flavor becomes visible.

[0060] Next, we will explain using an example where the relationship (similarity) between the first and second evaluation images is calculated using the chi-squared distance. In the case of a lemon juice beverage, the evaluation results of the first and second evaluation images were integrated using correspondence analysis, and Figure 5 shows the results as a 2D map based on the chi-squared distance. As shown in Figure 5, the emotion of "relaxation" was plotted near flavors such as "floral." This indicates that the emotion of "relaxation" has a high correlation (similarity) with flavors such as "floral." Similarly, the emotions of "awakening" and "refreshment" were plotted near flavors such as "citrus" and "sourness," and the emotion of "liked by children" was plotted near the flavor "sweet," indicating a high correlation (similarity) between these two. Note that for clarity, the chi-squared distance is not shown numerically in the plots in Figure 5, but it is certainly possible to include numerical values ​​as well.

[0061] <Evaluation system for products with flavor> By utilizing the above-described method for evaluating flavored products, an evaluation system for flavored products can be constructed. Figure 6 is a simplified diagram showing the configuration of the flavored product evaluation system 10. The flavor evaluation system 10 comprises an acquisition unit 11, an analysis / calculation unit 12, a storage unit 13, a network 14, an information processing terminal 15, and a printing device 16.

[0062] The acquisition unit 11 acquires responses representing the degree of the first evaluation image recalled for the product (first evaluation result) and responses representing the degree of the second evaluation image recalled for the product (second evaluation result). The acquisition method involves inputting the first and second evaluation result data output via an external network or external memory through an input unit (not shown). However, if connected to an external device via a network 14 (described later), the first and second evaluation result data measured via the network 14 can be acquired. The analysis and calculation unit 12 analyzes the responses representing the degree of the first evaluation image recalled for the product (first evaluation result) and the responses representing the degree of the second evaluation image recalled for the product (second evaluation result). Furthermore, it integrates and analyzes these two results, the first evaluation result and the second evaluation result, to calculate the relationship between the first evaluation image and the second evaluation image, etc. Note that the analysis and calculation unit 12 may be divided into two blocks that perform analysis and calculation separately (not shown).

[0063] The storage unit 13 stores at least the calculation results calculated by the analysis / calculation unit 12, and may also store other data calculated by the analysis / calculation unit 12. Furthermore, it is preferable that the calculation results stored in the storage unit 13 are updated as more analysis results input to the analysis / calculation unit 12 are accumulated, thereby improving the accuracy to the extent that the corresponding first evaluation image or second evaluation image can be recalled based on the calculation results.

[0064] The above process (shown in Figure 1) can be achieved by having the evaluation system 10 execute the program. Alternatively, the program can be recorded on an external recording medium, and the evaluation system 10 can read this external recording medium containing the program to achieve the above process. Various types of recording media can be used as the external recording medium, such as CD (Compact Disc)-ROM, flexible disk, magneto-optical disk, etc., which record information optically, electrically, or magnetically, or semiconductor memory such as ROM, flash memory, etc., which record information electrically. In the example shown in Figure 6, the functions of the acquisition unit 11, analysis / calculation unit 12, and storage unit 13 shown in Figure 6 can be realized by a control unit (not shown) executing one or more programs in the storage device. Furthermore, the storage unit 13 shown in Figure 6 can be realized by at least one or any combination of two or more storage devices and recording media.

[0065] Network 14 connects the storage unit 13 to external devices (for example, an information processing terminal 15 or a printing device 16) in a communicative manner, and plays a role in disseminating the data stored in the storage unit 13 over a wide area. Specifically, Network 14 consists of information and communication lines such as the Internet, a LAN (Local Area Network), and a mobile phone communication network, as well as the hardware that constructs these lines. Furthermore, the network 14 may be connected to devices other than the storage unit 13 and external devices. For example, in addition to the acquisition unit 11 described above, the analysis / calculation unit 12 may be connected to an external memory via the network 14 and perform faster analysis / calculation by using the external memory in conjunction with the network.

[0066] The information processing terminal 15 can acquire data stored in the storage unit 13 (including calculation results by the analysis / calculation unit 12) via the network 14 and display the acquired data. The printing device 16 can also acquire data stored in the storage unit 13 via the network 14 and print the acquired data. In other words, both the information processing terminal 15 and the printing device 16 can function as output units that output calculation results calculated by the analysis / calculation unit 12. The form of output by the information processing terminal 15 or the printing device 16 is not particularly limited. For example, a graph plotted in the first and second principal component spaces of a principal component analysis using a correlation matrix, as shown in Figure 4, may be output, or a table listing the respective correlation coefficients, as shown in Tables 2 and 3, may be output.

[0067] <Variation> The present invention is not limited to the embodiments described above, and the contents of the embodiments described above may be modified, improved, etc., insofar as the objectives of the present invention are achieved. For example, the first and second evaluation images may include not only language, but also some form of visual or auditory expression such as color, images, or sound. Furthermore, the correlation may be calculated by adding other different evaluation factors besides the first and second evaluation images.

[0068] Specifically, the second evaluation image is preferably an emotional image, as mentioned above, but emotional images can include not only language such as emotional words, but also visual expressions such as colors. Figure 7 shows a graph plotted in the first and second principal component spaces of the principal component analysis using a correlation matrix. Shampoo was selected as a product with a fragrance, and fragrance and emotional (color) evaluations were performed on four samples, A to D. The results of these evaluations were then analyzed, and principal component analysis was performed using a correlation matrix. In this case, five colors were extracted as emotional images. As is clear from the results shown in Figure 7, there are correlations between flavors such as "herbal" and "lily of the valley" and the emotion of "green," between flavors such as "citrus" and the emotion of "blue," between flavors such as "tropical" and "fruity" and the emotions of "yellow" and "orange," and between flavors such as "floral," "peach," "rose," and "berry" and the emotions of "red" and "pink."

[0069] Furthermore, Japanese Patent No. 6826195 discloses a method for representing color schemes inspired by fragrances as color representation diagrams. By incorporating these color schemes as evaluation factors, it becomes possible to plan and develop products with consistent concepts, flavors, and packaging during the planning and development stages of flavored products, and to imbue the products with corresponding flavors. By selling such products, it becomes possible to gain consumer empathy and strengthen the brand.

[0070] Furthermore, while emotions include visually evoked images and scenes in which one would want to consume the product, treating such images as separate evaluation factors, it is possible to further improve the matching of image with actual flavor, thereby broadening the reach and depth of sales for flavored products. Furthermore, it is possible to calculate the relationships between separate images included in the first evaluation image, and between separate images included in the second evaluation image, allowing for a deeper analysis of specific images based on these results. This analysis may also serve as a basis for developers to make proposals that better align with the concept.

[0071] Furthermore, in this invention, by making the second evaluation image the same as the first evaluation image (for example, if the first evaluation image is a flavor word, the second evaluation image is also a flavor word), it is possible to calculate the relationship between different subjects who have the same evaluation image. In other words, by having subjects with different attributes evaluate the same evaluation image, it becomes possible to clearly identify the differences in evaluations of that evaluation image among these different subjects (evaluation groups), and to effectively adjust for the differences in evaluation between the perspective of a developer with specialized knowledge and the perspective of the end user, the consumer.

[0072] Finally, by accumulating data using the evaluation methods and systems described above, it is possible to construct a system that can search, predict, and support development regarding the relationship between flavor and emotion. Such a system would allow for searching what emotions are strongly evoked by the flavors of existing products, and for predicting in advance what emotions are strongly evoked by the flavors of new products when planning and developing them. Furthermore, it would also be possible to provide support in determining the concept and flavor when developing new products.

[0073] Although the present invention has been described above using embodiments, it goes without saying that the technical scope of the present invention is not limited to the scope described in the above embodiments. It will be obvious to those skilled in the art that various modifications or improvements can be made to the above embodiments. Furthermore, it is clear from the claims that such modified or improved forms may also be included in the technical scope of the present invention. [Examples]

[0074] The present invention will be described in detail below with reference to examples. However, the present invention is not limited in any way to the examples shown below.

[0075] <Example 1: Ice cream> In Example 1, ice cream was selected as a product with flavor, and flavor and emotional evaluations were performed on four types of samples, A to D. The evaluation results for each were then analyzed, and principal component analysis was performed using a correlation matrix. Figure 8 shows graphs plotted in the first and second principal component spaces of the principal component analysis using the correlation matrix.

[0076] As is clear from the results shown in Figure 8, there is a correlation between the flavor "sweet aroma" and the emotion of "wanting to relax," between flavors such as "milky" and "vanilla (extract)" and the emotions of "luxury" and "wanting to keep it all to myself," between the flavor "crisp" and the emotion of "when it's hot," and between the flavor "custard" and the emotion of "classic / familiar."

[0077] Furthermore, Table 4 shows the correlation coefficients of each variable with respect to one of the two variables, while keeping the other variable fixed. [Table 4]

[0078] As shown in Table 4, emotions such as "classic / familiar," "reassuring," and "to be enjoyed with friends and family" were found to have a positive correlation with the flavor "custard (egg-like)" and a negative correlation with the flavor "vanilla extract." Furthermore, it was revealed that emotions such as "a sense of luxury / extravagance," "wanting to keep it all to myself," and "wanting to relax and unwind" have a strong positive correlation with flavors such as "vanilla extract," "milky richness," and "richness (body)," while simultaneously having a strong negative correlation with the flavor of "crispness." Furthermore, it was revealed that the sentiment of "when it's hot or after a bath" has a strong positive correlation with the flavor "crispness," and at the same time, a strong negative correlation with flavors such as "milky."

[0079] <Example 2: Fresh cream> In Example 2, fresh cream was selected as the product with flavor, and flavor and emotional evaluations were performed on five types of samples, A to E. The evaluation results for each were then analyzed, and principal component analysis was performed using a correlation matrix. Figure 9 shows graphs plotted in the first and second principal component spaces of the principal component analysis using the correlation matrix.

[0080] As is clear from the results shown in Figure 9, there is a correlation between flavors such as "fresh milk" and "Western liquor" and emotions such as "luxury / high-class" and "sophisticated," between flavors such as "sweet aroma" and "stickiness" and emotions such as "retro" and "classic / familiar," and between flavors such as "sweet" and "fruity" and emotions such as "likely to be liked by children."

[0081] <Example 3: Fragrance> In Example 3, fragrances were selected as products with flavor, and flavor and emotional evaluations were performed on five types of samples, A to E. After analyzing the evaluation results for each, principal component analysis was performed using a correlation matrix. Figure 10 shows graphs plotted in the first and second principal component spaces of the principal component analysis using the correlation matrix.

[0082] As is clear from the results shown in Figure 10, there is a correlation between flavors such as "tropical" and "sweet" and the emotion of "exotic," between flavors such as "sweet and sour" and the emotion of "invigorating," between flavors such as "fresh" and "crisp" and the emotions of "transparency" and "refreshment," and between flavors such as "gorgeous" and the emotion of "luxury / luxury."

[0083] <Example 4: Shampoo> In Example 4, shampoo was selected as a product with a fragrance, and fragrance and emotional evaluations were performed on four types of samples, A to D. After analyzing the evaluation results for each, principal component analysis was performed using a correlation matrix. Figure 11 shows graphs plotted in the first and second principal component spaces of the principal component analysis using the correlation matrix.

[0084] As is clear from the results shown in Figure 11, there is a correlation between flavors such as "herbal" and "green" and emotions such as "mature" and "natural," between flavors such as "apple" and "citrus" and emotions such as "refreshing" and "juicy," and between flavors such as "floral" and "peach" and emotions such as "youthful" and "cute."

[0085] Furthermore, Table 5 shows the correlation coefficients between emotions and the other variable, with one of the two variables fixed. [Table 5]

[0086] As shown in Table 5, it was revealed that the emotion of "cute" is positively correlated with emotions such as "glamorous," "youthful," and "energetic."

[0087] Table 6 shows the correlation coefficients between one of the two variables, flavor and emotion, and the other variable, with respect to that variable. [Table 6]

[0088] As shown in Table 6, it was revealed that the emotion of "cute" is positively correlated with aromas such as "rose," "floral," and "fruity."

[0089] Based on the above examples, it has been confirmed that the present invention provides an evaluation method and evaluation system that can visualize the relationship between what emotions are evoked by a particular flavor when the flavor of a product is ingested, and to what extent what flavors are evoked by a particular emotion. [Explanation of Symbols]

[0090] 10. Evaluation System 11 Acquisition Department 12 Analysis and calculation section 13 Storage Unit 14 Networks 15 Information Processing Terminals 16 Printing device

Claims

1. A product evaluation method that assesses the relationships between different evaluation images of the flavor and aroma of a product, A first evaluation step involves analyzing the first evaluation results, which represent the degree of the first evaluation image associated with the product using one or more numerical items. A second evaluation step involves analyzing the second evaluation results, which express the degree of the second evaluation image associated with the aforementioned product using one or more numerical items. A relationship calculation step involves integrating and analyzing the analysis results of the first evaluation result and the analysis results of the second evaluation result to calculate the relationship between the first evaluation image and the second evaluation image. An evaluation method characterized by comprising the following features.

2. The evaluation method according to claim 1, characterized in that the first subject evaluating the first evaluation result includes an expert panelist.

3. The evaluation method according to claim 1, characterized in that the first evaluation image includes a flavor image related to the flavor recalled by the first subject.

4. The evaluation method according to claim 1, characterized in that the second subject evaluating the second evaluation result includes general consumers.

5. The evaluation method according to claim 1, characterized in that the second evaluation image includes an emotional image relating to an emotion recalled by the second subject.

6. In the first evaluation step and the second evaluation step, the analysis method for the first evaluation result and the second evaluation result is principal component analysis or correspondence analysis. If the analysis method is principal component analysis, in the correlation calculation step, the correlation between the first evaluation image and the second evaluation image is calculated using a correlation matrix. The evaluation method according to claim 1, characterized in that, when the analysis method is correspondence analysis, the relationship between the first evaluation image and the second evaluation image is calculated as a chi-squared distance in the relationship calculation step.

7. A product evaluation method that assesses the relationship between aroma and emotion in the flavor of a product, An evaluation method characterized by comprising the step of analyzing a flavor evaluation representing the degree of the flavor image evoked by the product and an emotional evaluation representing the degree of the emotional image evoked by the product, and calculating the relationship between the flavor evaluation and the emotional evaluation.

8. A product evaluation system that evaluates the relationships between different evaluation images of the flavor and aroma of a product, An analysis unit analyzes a first evaluation result, which expresses the degree of the first evaluation image evoked for the product using one or more numerical items, and a second evaluation result, which expresses the degree of the second evaluation image evoked for the product using one or more numerical items. A calculation unit that integrates and analyzes the analysis results of the first evaluation result and the analysis results of the second evaluation result, and calculates the relationship between the first evaluation image and the second evaluation image, An evaluation system characterized by comprising the following features.

Images