Beauty evaluation system and sensory evaluation system

Abstract

PURPOSE: To provide a beauty determination system using an interview which can clarify a relation between a type of the skin of a client in relation to being oily or being dry and awareness of having a sensitive skin and allows more accurate determination of the skin type of the client.SOLUTION: The present invention relates to a beauty determination system for determining a beauty by processing interview data made of answers of a client to questions regarding the skin of the client by execution of a program, the beauty determination system including: storage means for storing a factor score coefficient; and skin type determination means for calculating a score SO of an O-factor regarding excessive skin oil of the client, a dry level SD as a score of a D-factor regarding a dry skin, and a score SH of an H-factor regarding a sense of pore-clogging debris by using the interview data and the factor score coefficient and determining the skin type of the client from positions of points (SO, SD, and SH) in a skin type identification space having an X-coordinate axis, a Y-coordinate axis, and a Z-coordinate axis. The factor score coefficient is calculated by extracting each of the factors from basic data made of answers of a large number of subjects to the questions regarding skins by a factor analysis technique.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a beauty judgment system for evaluating a customer's or subject's skin type based on questionnaire data, and a sensory evaluation system.

Background Art

[0002] As a technique for performing beauty judgment to evaluate a customer's or subject's skin type based on questionnaire data, for example, the techniques disclosed in Patent Documents 1 to 6 are known.

[0003] Patent Document 1 discloses a system for assigning a skin type to an individual in order to recommend or propose a skin care product, which uses a first element representing the degree of presence of oily or dry skin and a second element representing the degree of presence of sensitive or resistant skin. In this technique, the oiliness and dryness of the skin are relative concepts and not independent of each other (paragraph 0023 of Patent Document 1). Also, the relationship between the oiliness or dryness of the skin and sensitivity is not clear.

[0004] Patent Document 2 describes classifying the skin quality of a user into any one of dry skin, sensitive skin, normal skin, and oily skin based on a questionnaire, and conducting a questionnaire to ask about the skin sensitivity, but the details of the classification method are unclear.

[0005] Patent Document 3 states that there is no correlation between the lipid secretion characteristics and the water retention of the skin, and that dry skin and oily skin are not relative concepts but independent concepts. Further, this document describes providing four skin classifications of normal skin, oily skin, dry skin, and oily-dry skin using the lipid secretion status and the condition of roughness of the entire face and each part of the face as indicators, but there is no description about sensitive skin or skin sensitivity.

[0006] Patent Document 4 discloses a technique for classifying skin symptoms that arise from a customer's constitution according to traditional Chinese medicine theory into five types based on a medical interview. Of the five types, Type A is exemplified by symptoms such as oiliness, acne, and enlarged pores; Type B is exemplified by symptoms such as dryness, rough skin, and coarse texture; and Type E is exemplified by symptoms of sensitivity. However, the relationship between Type A (oily skin) and Type B (dry skin) and Type E (sensitive skin) is not clear.

[0007] Patent Document 5 discloses a technique that, based on the principles of traditional Chinese medicine, uses multi-class linear discriminant analysis on questionnaire responses to classify a user's skin composition into one of five distinct categories: Yang-dominant, Balanced-Yang, Balanced, Balanced-Yin, and Yin-dominant. Here, "Yang-dominant" includes oily skin and "Yin-dominant" includes dry skin; therefore, in this technique, "oily skin" and "dry skin" are not considered independent but rather opposing concepts. Furthermore, it is unclear where sensitive skin fits into these five classifications.

[0008] Patent Document 6 discloses a technique for collecting basic data by conducting interviews with subjects of different ages, extracting factors for excessive sebum and skin dryness using factor analysis, and then classifying the skin type of a subject in two dimensions based on the interview data when interviewing a subject of a specific age. However, with this technique, it is unclear where sensitive skin is positioned in the above two-dimensional classification.

[0009] Many women identify as having "sensitive skin," and cosmetics for sensitive skin are available, but the meaning of "sensitive skin" (hereinafter referred to as "sensitive skin awareness") is not always clear. Generally, "sensitive skin" refers to hypersensitivity to stimuli such as chemical substances, sunlight (ultraviolet rays), pressure, friction, and temperature, and does not necessarily reflect the condition of the subject's skin or their physical and mental health. It is thought to reflect the state of being, etc. However, for example, the relationship between the subject's skin condition (skin type related to oiliness or dryness) and their self-perception of sensitive skin is not clear at all. [Prior art documents] [Patent Documents]

[0010] [Patent Document 1] Special Publication No. 2008-521145 [Patent Document 2] Japanese Patent Publication No. 2002-251532 [Patent Document 3] Japanese Patent Publication No. 2001-275990 [Patent Document 4] Japanese Patent Publication No. 2002-140434 [Patent Document 5] Patent No. 5400217 [Patent Document 6] Patent No. 5419004 [Overview of the project] [Problems that the invention aims to solve]

[0011] One aspect of the present invention aims to provide a beauty assessment system that can more accurately determine the skin type of customers who perceive themselves as having sensitive skin, by clarifying the relationship between oily or dry skin type and self-perceived sensitive skin in a large number of subjects. Another aspect of the present invention aims to provide a sensory evaluation system for beauty products using the above-described beauty assessment system. [Means for solving the problem]

[0012] The present invention has been made to solve such problems. The first aspect of the present invention is a beauty determination system that consists of an independent computer or a plurality of computers capable of two-way communication with each other, and processes questionnaire data, which is the customer's answers to three or more questions regarding the skin, by executing a computer control program to perform the beauty determination of the customer. The system has storage means for storing the customer's questionnaire data, a computer control program, and factor score coefficients. The computer control program uses the customer's questionnaire data and the factor score coefficients to calculate the score SO of the O factor related to sebum excess, the dryness SD which is the score of the D factor related to skin dryness, and the score SH of the H factor related to the feeling of pore dirtiness of the customer, and has skin type determination means for determining the customer's skin type from the position of the point (SO, SD, SH) in the skin type identification space equipped with XYZ coordinate axes. The factor score coefficients are calculated in advance by factor analysis from basic data which is the answers of a large number of subjects to the questions regarding the skin, and at least extract the O factor, the D factor, and the H factor which are linearly independent of each other.

[0013] The second aspect of the present invention is the beauty determination system according to the first aspect, wherein the questions include at least one of the items regarding pores, pore dirt, pore opening, and pore darkening, which are highly related to the feeling of pore dirtiness.

[0014] The third aspect of the present invention is the beauty determination system according to the second aspect, wherein using the score SO of the O factor, the score SH of the H factor, and an angle t where 0° < t < 90°, the composite lipidicity SC is determined by the following formula 1: (Formula 1) SC = SO×cos(t) + SH×sin(t) The skin type determination means calculates the composite lipidicity SC of the customer based on the questionnaire data, and determines the customer's skin type from the position of the point (SC, SD) in the skin type identification plane equipped with XY coordinate axes. In the following specification text, tables, or drawings, "composite lipidicity" may be simply referred to as "lipidicity", but unless otherwise specified, the two are synonymous.

[0015] The fourth aspect of the present invention is, in the third aspect, the computer control program uses the composite lipid degree SC, dryness degree SD, and an angle s where 0° < s < 90° of a customer or a subject to have a means for calculating the skin quality score SQ of the customer or the subject by the following formula (2), (Formula 2) SQ = SC × cos(s) + SD × sin(s) The question regarding the basic data includes a question about whether the subject is aware that their skin type is sensitive skin, which is a question about the presence or absence of awareness of sensitive skin, separate from the questions regarding the answers used in the factor analysis. The beauty determination system is such that the angle t and the angle s are determined so that the correlation coefficient R calculated from the distribution of the skin quality score SQ of the subject and the answer to the question about the presence or absence of awareness of sensitive skin in the basic data is 0.90 or more.

[0016] The fifth aspect of the present invention is, in the fourth aspect, the interview data further includes the customer's answers to a plurality of questions regarding eating habits, the storage means further stores factor score coefficients related to eating habits, and the computer control program uses the factor score coefficients related to eating habits and the customer's interview data to calculate the score SF1 of the F1 factor related to irregular eating and the score SF2 of the F2 factor related to unhealthy eating of the customer. Further, using an angle u where 0° < u < 90°, it has a means for calculating the score SF of the composite diet factor of the customer by the following formula (3), (Formula 3) SF = SF1 × cos(u) + SF2 × sin(u) The basic data further includes dietary lifestyle basic data which is the answers of a large number of the subjects to the plurality of questions regarding dietary lifestyle. The factor score coefficients related to dietary lifestyle are calculated in advance by extracting, from the dietary lifestyle basic data by factor analysis, at least an F1 factor related to irregular eating and an F2 factor related to unhealthy eating, which are linearly independent of each other. Using the factor score coefficients related to dietary lifestyle, the score SF1 of the F1 factor and the score SF2 of the F2 factor of each subject in the dietary lifestyle basic data are calculated in advance. Further, the score SF of the composite diet factor is calculated in advance using Equation 3. The beauty determination system is such that the angle u is determined so that the correlation coefficient R calculated from the distribution of the score SF of the composite diet factor of the subject and the answer to the question about the presence or absence of self-awareness of sensitive skin in the basic data is 0.90 or more.

[0017] In a sixth aspect of the present invention, in the fourth aspect, the interview data further includes the customer's answers to a plurality of questions regarding physical condition awareness. The storage means further stores factor score coefficients related to physical condition awareness. The computer control program has means for calculating the score ST of a T factor related to the physical condition specific to menopause of the customer using the factor score coefficients related to physical condition awareness and the customer's interview data. The basic data further includes health basic data which is the answers of a large number of the subjects to the plurality of questions regarding physical condition awareness. The factor score coefficients related to physical condition awareness are calculated in advance by extracting, from the health basic data by factor analysis, at least an A factor related to the menopause age and a T factor related to the physical condition specific to menopause, which are linearly independent of each other. Using the factor score coefficients related to physical condition awareness, the score ST of the T factor of each subject in the health basic data is calculated in advance. The beauty determination system is such that the correlation coefficient R calculated from the distribution of the score ST of the T factor of the subject and the answer to the question about the presence or absence of self-awareness of sensitive skin in the basic data is 0.85 or more.

[0018] The seventh aspect of the present invention is, in the sixth aspect, that the computer control program has means for calculating a score SK of a circulation factor related to circulatory disorder of the customer using the factor score coefficient related to physical condition awareness and the interview data of the customer, the factor score coefficient related to physical condition awareness is calculated by extracting at least a circulation factor related to circulatory disorder from the health basic data by factor analysis, using the factor score coefficient related to physical condition awareness, the score SK of the circulation factor of each subject in the health basic data has been calculated in advance, and the correlation coefficient R calculated from the distribution of the score SK of the circulation factor of the subject and the answer to the question of whether or not there is a feeling of sensitive skin in the basic data is 0.90 or more.

[0019] The eighth aspect of the present invention is, in the sixth aspect, that the computer control program has means for calculating a score SAL of an AL factor related to allergy of the customer using the factor score coefficient related to physical condition awareness and the interview data of the customer, the factor score coefficient related to physical condition awareness is calculated by extracting at least an AL factor related to allergy from the health basic data by factor analysis, using the factor score coefficient related to physical condition awareness, the score SAL of the AL factor of each subject in the health basic data has been calculated in advance, and the correlation coefficient R calculated from the distribution of the score SAL of the AL factor of the subject and the answer to the question of whether or not there is a feeling of sensitive skin in the basic data is 0.85 or more.

[0020] The ninth aspect of the present invention is a beauty determination system in any one of the fourth to eighth aspects, wherein the angle t is 25° or more and 50° or less, and the angle s is 25° or more and 75° or less.

[0021] The tenth aspect of the present invention is the computer control program in the ninth aspect.

[0022] An eleventh embodiment of the present invention is a computer-readable storage medium that records parameters related to the computer control program in the ninth embodiment.

[0023] A twelfth embodiment of the present invention is a sensory evaluation system for beauty products that utilizes the beauty evaluation system of the ninth embodiment, wherein a collaborator who cooperates in the evaluation uses the beauty product for a certain period of time, and the collaborator is interviewed both before and after the period, and based on the collaborator's answers to the questions, one or more numerical values ​​from among the O factor score, H factor score, combined oiliness, dryness, and skin quality score are calculated both before and after the period, and the improvement effect is evaluated from the amount of change in the numerical values ​​during the period. [Effects of the Invention]

[0024] According to one embodiment of the present invention, a beauty assessment system can be provided that can more accurately determine the skin type of customers who perceive themselves as having sensitive skin by clarifying the relationship between the oily or dry skin type and the perception of sensitive skin among a large number of test subjects. According to another embodiment of the present invention, a sensory evaluation system for beauty products can be provided using the above-mentioned beauty assessment system. [Brief explanation of the drawing]

[0025] [Figure 1] Figure 1A is an explanatory diagram showing the geometric meaning of the combined oiliness score SC and angle t, and Figure 1B is an explanatory diagram showing the geometric meaning of the skin quality score SQ and angle s. [Figure 2] Figures 2A and 2B are tables showing the questions related to beauty. [Figure 3] Figures 3A and 3B are tables showing the factor weightings and factor score coefficients related to beauty. [Figure 4-1] Figures 4-1A and 4-1B are graphs (contour plots) showing the distribution of the proportion of people who perceive themselves as having sensitive skin within the skin type identification space. [Figure 4-2]Figures 4-2A and 4-2B are graphs showing the distribution of the percentage of people who perceive themselves as having oily skin. [Figure 4-3] Figures 4-3A and 4-3B are graphs showing the distribution of the percentage of people who perceive themselves as having dry skin. [Figure 4-4] Figures 4-4A and 4-4B are graphs showing the distribution of the percentage of people who perceive themselves as having normal skin. [Figure 4-5] Figures 4-5A and 4-5B are graphs showing the distribution of the percentage of people who perceive themselves as having combination skin. [Figure 4-6] Figures 4-6A and 4-6B are graphs (contour plots) showing the distribution of the proportion of individuals who perceive themselves as having irregular or unhealthy eating habits within the skin type identification space. [Figure 4-7] Figures 4-7A and 4-7B are graphs (contour plots) showing the distribution of the proportion of people who self-report circulatory problems in the skin type identification space. [Figure 5-1] Figures 5A, 5B, and 5C are tables showing the distribution of the proportion of individuals who perceive themselves as having various skin types on the skin type identification plane. [Figure 5-2] Figures 5D, 5E, and 5F are tables showing the distribution of the proportion of individuals who perceive themselves as having various skin types on the skin type identification plane, as well as the distribution of test subjects. [Figure 6-1] Figure 6-1 is a graph (contour plot) showing the distribution of the percentage of people who perceive themselves as having sensitive skin on a skin type identification plane. [Figure 6-2] Figure 6-2 is a graph showing the distribution of the percentage of people who perceive themselves as having oily skin. [Figure 6-3] Figure 6-3 is a graph showing the distribution of the percentage of people who perceive themselves as having dry skin. [Figure 6-4] Figure 6-4 is a graph showing the distribution of the percentage of people who perceive themselves as having normal skin. [Figure 6-5] Figure 6-5 is a graph showing the distribution of the percentage of people who perceive themselves as having combination skin. [Figure 6-6] Figure 6-6 is a graph (contour plot) showing the distribution of the proportion of people who perceive themselves as having irregular or unhealthy eating habits on the skin type identification plane. [Figure 6-7]Figure 6-7 is a graph (contour plot) showing the distribution of the percentage of people who self-report circulatory problems on the skin type identification plane. [Figure 7] Figure 7 is a table showing how the correlation coefficient R, calculated from the distribution of skin type score SQ and responses to the question of whether or not one perceives oneself as having sensitive skin, changes with angles t and s. [Figure 8] Figure 8 is a table showing the questions related to dietary habits. [Figure 9] Figure 9 is a table showing the factor score coefficients for dietary habits. [Figure 10] Figure 10 is a table showing the questionnaire items regarding self-perception of physical condition. [Figure 11] Figure 11 is a table showing the factor score coefficients for subjective perception of physical condition. [Figure 12] Figure 12A is a diagram of the beauty evaluation system consisting of independent computers, and Figure 12B is a diagram of the computer configuration. [Figure 13] Figure 13 is a diagram showing the configuration of a computer control program. [Figure 14] Figure 14A is a configuration diagram showing an example of a beauty evaluation system consisting of a server and terminals that can communicate via a network, and Figure 14B is a configuration diagram showing another example of a similar configuration. [Figure 15A] Figure 15A is a graph showing the relationship between skin type score and self-perception of sensitive skin. [Figure 15B] Figure 15B is a graph showing the relationship between irregular eating habits and self-perceived sensitive skin. [Figure 15C] Figure 15C is a graph showing the relationship between an unhealthy diet and the perception of sensitive skin. [Figure 15D] Figure 15D is a graph showing the relationship between scores on the complex diet factor and self-perception of sensitive skin. [Figure 15E] Figure 15E is a graph showing the relationship between the score for the menopausal age factor and the self-perceived sensitive skin. [Figure 15F] Figure 15F is a graph showing the relationship between scores on menopausal-specific physical factors and self-reported sensitive skin. [Figure 15G]Figure 15G is a graph showing the relationship between the score for circulating factors and the self-perception of sensitive skin. [Figure 15H] Figure 15H is a graph showing the relationship between allergy factor scores and self-perceived sensitive skin. [Modes for carrying out the invention]

[0026] Next, embodiments of the beauty evaluation system and sensory evaluation system according to the present invention will be described in detail with reference to the drawings.

[0027] According to a first embodiment of the present invention, a beauty assessment system comprising an independent computer or a plurality of computers capable of bidirectional communication with each other processes questionnaire data, which is the customer's answers to three or more questions about skin, by executing a computer control program to perform a beauty assessment of the customer, and has storage means for storing the customer's questionnaire data, the computer control program, and factor score coefficients, wherein the computer control program uses the customer's questionnaire data and the factor score coefficients to determine the customer's O factor related to excessive sebum The system provides a beauty determination system that calculates a score SO, a dryness SD which is the score for factor D related to skin dryness, and a score SH for factor H related to the sensation of clogged pores, and determines the customer's skin type from the position of a point (SO, SD, SH) in a skin type identification space equipped with XYZ coordinate axes, wherein the factor score coefficients are pre-calculated by extracting at least the O factor, the D factor, and the H factor, which are linearly independent of each other (more preferably orthogonal to each other), from basic data consisting of the answers of many subjects to the aforementioned questions about skin, using a factor analysis method.

[0028] (Advantages of using factor analysis when determining skin type through medical interviews) Surprisingly, customers' subjective perceptions of skin dryness and oiliness have a low correlation with measurements taken by sebum and moisture levels using measuring instruments. This is thought to be related to the fact that the environment of the measurement location, such as heating and air conditioning, affects skin condition, and that measurements are often taken with makeup on or immediately after washing the face. In this embodiment of the present invention, when determining skin type through interviews, instead of determining skin type solely from the customer's answers to direct questions regarding excess sebum, pore dirt sensation, and skin dryness, the number of questions directly asking about the degree of skin dryness and oiliness is reduced, and the scores for each factor of the customer are calculated by comprehensively analyzing the answers to all questions using factor analysis. The customer's skin type is then determined comprehensively and indirectly based on these factor scores, thereby improving the accuracy of the determination. In addition, since the customer's skin type is accurately determined from the position of points (SO, SD, SH) in a skin type identification space equipped with XYZ coordinate axes, both the customer and the seller can quickly grasp and share information about the customer's skin type.

[0029] (Number of subjects in the basic data) Here, "a large number of subjects" in the basic data refers to a number of subjects that is not limited to 1,000 or more, preferably 5,000 or more, and even more preferably 10,000 or more. The larger the basic data, which consists of responses from subjects belonging to a sample group randomly sampled from the target customer population, the smaller the discrepancy between the relative position of each customer in the sample group, as grasped by the scores of each factor, and the true relative position of that customer in the population, and the more accurate the skin type determination becomes.

[0030] (The impact of the size of the baseline data on the extraction of factors O and H) As a guideline for selecting the number of factors in factor analysis, for example, a method (Kaiser - Guttman criterion) of setting the number of eigenvalues greater than 1 in the sample correlation matrix as the number of factors is well - known. According to the inventor's experience, the O factor related to sebum excess and the H factor related to the feeling of clogged pores cannot be separated and extracted as two linearly independent factors with explanatory power for the answers to the questions when the number of subjects included in the basic data for factor analysis is about 1,000. When the number of subjects is about 5,000, they can barely be separated and extracted, and when the number of subjects is about 10,000, they can surely be separated and extracted. In the prior art described in Patent Document 6 by the inventor, since the number of subjects included in the basic data was 4,227, the O factor and the H factor could not be extracted as two independent factors, and both were grasped as one factor related to fatness with an overlap. In one embodiment of the present invention, since the number of subjects included in the basic data is as large as about 20,000, the O factor and the H factor can be extracted as two independent factors, so that the customer's skin type can be determined more precisely and in finer detail.

[0031] According to a second aspect of the present invention, in the first aspect, a beauty determination system can be provided in which the questionnaire includes at least one item related to pores, pore dirt, pore opening, and pore darkening, which are closely related to the feeling of clogged pores.

[0032] According to a third aspect of the present invention, in the second aspect, using the score SO of the O factor, the score SH of the H factor, and an angle t such that 0° < t < 90°, the composite fatness degree SC is determined by the following formula 1, and (Formula 1) SC = SO×cos(t) + SH×sin(t) The skin type determination means can provide a beauty determination system that calculates the composite fatness degree SC of the customer based on the interview data and determines the customer's skin type from the position of the point (SC, SD) in the skin type identification plane having XY coordinate axes.

[0033] (Graphical meaning of the composite fatness degree SC) Explain the graphical meaning of the composite lipidiness SC defined by Equation 1. As shown in Figure 1A, in the coordinate plane with the score SO of the O factor on the horizontal axis (referred to as the O axis) and the score SH of the H factor on the vertical axis (referred to as the H axis), draw a straight line passing through the origin and making an angle t with the horizontal axis in the first quadrant, and call this straight line the C axis. Let the foot of the perpendicular from the point (SO, SH) to the C axis be point A. The distance between point A and the origin (more precisely, the signed distance) is the composite lipidiness SC. (In Figure 1A, "composite lipidiness" is simply denoted as "lipidiness".)

[0034] (Motivation for defining the composite lipidiness SC) As described above, the O factor related to sebum excess and the H factor related to the feeling of clogged pores are linearly independent of each other. However, the scores of these factors are both high in the young female group in their 20s and tend to be low in non-young female groups, which is consistent with the general lipid feeling of women (the lipid feeling decreases as age progresses). Therefore, when the composite lipidiness SC that reflects both the score SO of the O factor and the score SH of the H factor is defined by Equation 1, this composite lipidiness SC can be used as an index representing the applicable lipidiness for women of all age groups.

[0035] In the present embodiment of the invention, since the composite lipidiness SC defined by Equation 1 and the dryness SD are used to determine the skin type of the customer, the skin type can be accurately determined for women of all age groups. Also, in the present embodiment of the invention, since the skin type of the customer is visually represented as the position of the point (SC, SD) in the skin type discrimination plane equipped with the XY coordinate axes, both the customer and the seller etc. can grasp and share the information on the skin type of the customer in a short time. Also, the customer and / or the seller etc. can grasp the skin type of the customer more detailedly and accurately and use it as a reference for the purchase and recommendation of beauty products.

[0036] According to the fourth embodiment of the present invention, in the third embodiment, the computer control program has means for calculating the skin quality score SQ of the customer or the subject using the composite lipidiness SC and the dryness SD of the customer or the subject and an angle s such that 0° < s < 90° according to the following Equation 2. (Equation 2) SQ = SC × cos(s) + SD × sin(s) The questions relating to the basic data include, separately from the questions relating to the answers used in the factor analysis, a question asking whether the subject is aware of having sensitive skin, and the angles t and s are set such that the correlation coefficient R calculated from the distribution of the subject's skin type score SQ and the answers to the question on whether they are aware of having sensitive skin in the basic data is 0.90 or higher.

[0037] (The geometric meaning of the skin quality score SQ) The geometric meaning of the skin quality score SQ, defined by Equation 2, is explained below. As shown in Figure 1B, in a coordinate plane with the combined oiliness SC on the horizontal axis (called the C axis) and the dryness SD on the vertical axis (called the D axis), a line passing through the origin and making an angle s with the horizontal axis in the first quadrant is taken and called the Q axis. Let B be the foot of the perpendicular from point (SC,SD) to the Q axis. The distance (more precisely, the signed distance) between point B and the origin is the skin quality score SQ. The angle s may or may not be equal to 45°. According to Equation 2, when the angle s is 45°, the combined oiliness SC and dryness SD contribute to the skin quality score SQ with the same weight; when the angle s is less than 45°, the combined oiliness SC contributes to the skin quality score SQ with a greater weight than dryness SD; and when the angle s is greater than 45°, dryness SD contributes to the skin quality score SQ with a greater weight than the combined oiliness SC. To side with.

[0038] (Answers to the question regarding whether or not you perceive yourself as having sensitive skin, and selection of angles t and s) In this embodiment of the present invention, the angles t in Equation 1 and s in Equation 2 are determined such that the correlation coefficient R calculated from the distribution of the subject's skin quality score SQ and the response to the question regarding whether or not they perceive themselves as having sensitive skin in the basic data is 0.90 or higher. Therefore, the skin type of customers who perceive themselves as having sensitive skin and those who do not can be more accurately grasped using the skin quality score SQ, which has a large correlation with the perception of sensitive skin, and the dryness degree SD, and this can be used as a reference for purchasing or recommending beauty products.

[0039] (Calculation of the correlation coefficient R) Here, the correlation coefficient R is calculated from the distribution of the subjects' skin type score SQ and their responses to the question about whether they perceive themselves as having sensitive skin, as follows: Let L be an integer of 3 or more, and for the subjects included in the basic data, consider L intervals from interval 1 to interval L to which the value of the skin type score SQ belongs [x0,x1), [x1,x2), ..., [x_(L-1), xL] (where x0=min(SQ), xL=max(SQ)). Here, the endpoints x0, x1, ..., xL of each interval are equally spaced, and preferably each interval contains at least one value of the subject's skin type score SQ. Let i be any integer between 1 and L, and let pi be the proportion of subjects who perceive themselves as having sensitive skin among the subjects whose skin type score SQ belongs to the i-th interval. Then calculate the correlation coefficient R of L points (i,pi) (i=1,2, ...,L) on the coordinate plane. This R is the correlation coefficient to be obtained, and in this embodiment of the present invention, the angles t and s are determined such that R is 0.90 or greater. Unless there is a special reason, the number of intervals is L=10 in principle. Furthermore, if the number of subjects whose skin quality score SQ value belongs to a certain interval (let's call it the kth interval) among the L intervals is less than a predetermined lower limit of the number of people per interval, in order to ensure the accuracy of the correlation coefficient R, the correlation coefficient R is calculated using the remaining points after removing the point (k,pk) corresponding to that interval (the kth interval) from L points (i,pi) (i=1,2,··,L) on the coordinate plane (if there are multiple points to be removed, all of those points are removed and the remaining points are used). Unless there is a special reason, the predetermined lower limit of the number of people per interval is the greater of 10 people and 1 / 100th of the number of valid respondents (for example, 64 people). Here, we have explained how to calculate the correlation coefficient R using the subject's skin type score SQ. However, instead of the skin type score SQ, any arbitrary quantity (real number), such as some factor score related to the subject, can be used, and the correlation coefficient calculated from that quantity and the distribution of responses to the question about whether or not the subject perceives sensitive skin can be calculated in the same manner.

[0040] According to the fifth aspect of the present invention, in the fourth aspect, the interview data further includes the customer's answers to a plurality of questions regarding diet, the storage means further stores a factor score coefficient related to diet, and the computer control program uses the factor score coefficient related to diet and the customer's interview data to calculate the score SF1 of the F1 factor related to irregular eating and the score SF2 of the F2 factor related to unhealthy eating of the customer, and further has means for calculating the score SF of the composite diet factor of the customer by the following formula 3 using an angle u such that 0° < u < 90°, (Formula 3) SF = SF1×cos(u)+ SF2×sin(u) The basic data further includes diet basic data which is the answers of a large number of the subjects to the plurality of questions regarding diet. The factor score coefficient related to diet is calculated in advance by extracting at least the F1 factor related to irregular eating and the F2 factor related to unhealthy eating, which are linearly independent of each other, from the diet basic data by factor analysis. Using the factor score coefficient related to diet, the score SF1 of the F1 factor and the score SF2 of the F2 factor of each subject in the diet basic data are calculated in advance, and further, the score SF of the composite diet factor is calculated in advance using Formula 3. The correlation coefficient R calculated from the distribution of the score SF of the composite diet factor of the subject and the answer to the question about the presence or absence of sensitive skin awareness in the basic data is 0.90 or more, so that a beauty determination system can be provided in which the angle u is determined. A beauty determination system can be provided in which the angle u is determined so that the correlation coefficient R calculated from the distribution of the score SF of the composite diet factor of the subject and the answer to the question about the presence or absence of sensitive skin awareness in the basic data is 0.90 or more.

[0041] (Relationship between the diet of the subject and the awareness of sensitive skin) The inventor has found that there is a large correlation between the score SF of the composite diet factor of the subject and the presence or absence of sensitive skin awareness. In this aspect of the present invention, customers and / or sellers, etc. can more accurately grasp the dietary patterns of customers with and without awareness of sensitive skin using the score SF1 of the F1 factor related to irregular eating, the score SF2 of the F2 factor related to unhealthy eating, and the score SF of the composite diet factor having a large correlation with sensitive skin awareness of the customer, and use it as a reference for the purchase and recommendation of beauty products and advice on diet.

[0042] According to a sixth embodiment of the present invention, in the fourth embodiment, the questionnaire data further includes the customer's answers to a plurality of questions regarding self-perceived physical condition, the storage means further stores factor score coefficients relating to self-perceived physical condition, the computer control program has means for calculating the customer's score ST of the T factor relating to menopausal physical condition using the factor score coefficient relating to self-perceived physical condition and the customer's questionnaire data, and the basic data further includes basic health data which is the answers of a large number of subjects to the plurality of questions regarding self-perceived physical condition, and the factor score coefficient relating to self-perceived physical condition The number is calculated in advance by extracting at least two linearly independent (more preferably orthogonal to each other) factors A relating to menopausal age and two factors T relating to physical condition specific to menopause from the aforementioned basic health data using factor analysis. The score ST of the T factor for each subject in the aforementioned basic health data is calculated in advance using the factor score coefficient relating to subjective perception of physical condition. A beauty judgment system can be provided in which the correlation coefficient R calculated from the distribution of the subject's T factor score ST in the aforementioned basic data and the responses to the question of whether or not they perceive sensitive skin is 0.85 or higher.

[0043] (The relationship between the subjects' self-awareness of menopausal-specific physical conditions and their self-awareness of sensitive skin) The inventors have found a significant correlation between the subject's T factor score (ST) related to menopausal physical condition and whether or not they perceive themselves as having sensitive skin. In this embodiment of the present invention, customers and / or sellers can more accurately grasp the physical condition perception of customers who perceive themselves as having sensitive skin and those who do not, using the customer's T factor score (ST) related to menopausal physical condition, and use this information as a reference for purchasing and recommending beauty products, as well as for providing health advice.

[0044] According to a seventh embodiment of the present invention, in the sixth embodiment, the computer control program has means for calculating the score SK of the circulatory factors related to circulatory disorders for the customer using the factor score coefficient related to the perception of physical condition and the customer's medical interview data, wherein the factor score coefficient related to the perception of physical condition is calculated by extracting at least the circulatory factors related to circulatory disorders from the basic health data using a factor analysis method, and the score SK of the circulatory factors for each subject in the basic health data is calculated in advance using the factor score coefficient related to the perception of physical condition, and a beauty judgment system can be provided in which the correlation coefficient R calculated from the distribution of the score SK of the circulatory factors for the subject in the basic data and the answers to the question of whether or not the subject perceives sensitive skin is 0.90 or higher.

[0045] (Relationship between subjects' scores on circulatory factors and their self-perception of sensitive skin) The inventors have found a significant correlation between the subject's score SK for the circulatory factors related to circulatory dysfunction and whether or not they perceive themselves as having sensitive skin. In this embodiment of the present invention, customers and / or sellers can more accurately grasp the physical condition perception of customers who perceive themselves as having sensitive skin and those who do not, using the customer's score SK for the circulatory factors, and use this information as a reference for purchasing and recommending beauty products, as well as for providing health advice.

[0046] According to an eighth embodiment of the present invention, in the sixth embodiment, the computer control program has means for calculating the customer's score SAL of the AL factor related to allergies using the factor score coefficient related to self-awareness of physical condition and the customer's medical interview data, wherein the factor score coefficient related to self-awareness of physical condition is calculated by extracting at least the AL factor related to allergies from the basic health data using a factor analysis method, and the score SAL of the AL factor for each subject in the basic health data is calculated in advance using the factor score coefficient related to self-awareness of physical condition, and a beauty judgment system can be provided in which the correlation coefficient R calculated from the distribution of the subject's score SAL of the AL factor in the basic data and the answers to the question about whether or not they are aware of sensitive skin is 0.85 or higher.

[0047] (Relationship between the subject's AL factor score and their self-perception of sensitive skin) The inventors have found a significant correlation between the subject's score on the AL factor related to allergies (SAL) and whether or not they perceive themselves as having sensitive skin. In this embodiment of the present invention, customers and / or sellers can more accurately grasp the physical condition perception of customers who perceive themselves as having sensitive skin and those who do not, using the customer's score on the AL factor (SAL), and use this information as a reference for purchasing and recommending beauty products, as well as for providing health advice.

[0048] According to the ninth embodiment of the present invention, in any of the fourth to eighth embodiments, a beauty assessment system can be provided in which the angle t is 25° or more and 50° or less, and the angle s is 25° or more and 75° or less. By selecting angles t and s to be within this range, a beauty assessment system is provided in which the correlation coefficient R calculated from the distribution of the subject's skin quality score SQ and the responses to the question of whether or not they perceive themselves as having sensitive skin in the basic data is 0.98 or more. Angle s may or may not be equal to 45°. In the latter case, angle s may be less than 45° or greater than 45°. Angle t may or may not be equal to 45°. In this embodiment, it can be said that the higher the customer's skin quality score SQ, the greater the probability that the customer perceives themselves as having sensitive skin.

[0049] According to a tenth embodiment of the present invention, the computer control program described in the ninth embodiment can be provided.

[0050] According to an eleventh embodiment of the present invention, a computer-readable storage medium can be provided that records parameters relating to the computer control program in the ninth embodiment. Here, the parameters relating to the computer control program are not limited to, but include, for example, factor weighting coefficients and factor score coefficients relating to the factor analysis of skin diagnosis, diet, or health, and the angles t, s, and u.

[0051] According to a twelfth embodiment of the present invention, a sensory evaluation system for a beauty product is provided that utilizes the beauty judgment system of the ninth embodiment, wherein a collaborator who cooperates in the evaluation uses the beauty product for a certain period of time, and interviews are conducted with the collaborator both before and after the use of the beauty product. Based on the collaborator's answers to the interview, one or more numerical values ​​from among the O factor score, H factor score, combined oiliness, dryness, and skin quality score are calculated for both before and after the period, and an improvement effect evaluation means is provided to evaluate the improvement effect from the amount of change in the numerical values ​​during the period.

[0052] Typically, in sensory evaluations conducted through interviews during the development phase of cosmetics, for example, taking the degree of "moisture" as an example, subjects are asked to test use the cosmetic product for a short period in a laboratory, and interviews are conducted before and after the test using direct questions. The selected evaluation options are used as a moisturizing scale, and the improvement effect is measured by the change in the moisturizing scale before and after the test. On the other hand, in this embodiment of the present invention, for example, a dryness SD score, which is a score of the dryness factor calculated from answers to questions that are not necessarily direct, is used over a certain period. The improvement effect is measured as a change in score before and after (for example, several days or more). This embodiment is a sensory evaluation method that does not use measuring instruments and can provide a sensory evaluation method that can be used in the actual usage situation of cosmetics or beauty services, in accordance with actual usage conditions outside the laboratory.

[0053] <1. A beauty assessment system that determines skin type using three factors> (Extraction of basic data and factors) The inventors interviewed 20,000 female subjects aged 29 to 69 using a set of 21 questions useful for skin type analysis, as shown in the table in Figure 2A, and obtained valid responses from 19,468 subjects. The results of the analysis performed using factor analysis on these valid responses as basic data, and a beauty assessment system according to one embodiment of the present invention will be explained. Figure 3A is a table showing the factor weights of five factors extracted by analyzing the basic data consisting of the two-choice answer values ​​(0 or 1) of each subject for each of the 21 questions in the above-mentioned set of questions, using factor analysis with an orthogonal model. The number of factors extracted was set to 5 because the number of eigenvalues ​​greater than or equal to 1 in the covariance matrix of the above answer values, standardized to mean 0 and standard deviation 1, is 5. The maximum likelihood method was used to estimate the factor weights. The factor weights shown in Figure 3A are the factor weights after six rotations using the varimax method without Kaiser normalization until convergence. Figure 3B is a table showing the factor score coefficients obtained from the analysis using the above factor analysis method. The regression method was used to calculate the factor scores. For any given factor, the factor score coefficient can be multiplied by the answer value for each subject to the question, and the sum is taken for all 21 questions to calculate the factor score for that subject (excluding constant deviations). Factor 1 is the factor that provides the highest explanatory power for the response value and relates to beauty perceptions such as awareness of skin whitening, and is strongly correlated with questions related to aesthetics (firmness and elasticity, brightness and transparency, skin texture, youthfulness). In this specification, this factor is called the EST factor. Factor 2 is the factor related to the perception of sebum (excess), and is strongly correlated with questions related to excessive sebum (acne, acne scars). In this specification, this factor is called the O factor, and its factor score is denoted as SO. Factor 3 is the factor related to the perception of pores (dirt), and is strongly correlated with questions related to pore dissatisfaction (pores, dirt in pores). In this specification, this factor is called the H factor, and its factor score is denoted as SH. Factor 4 is the factor related to the perception of base makeup, and is strongly correlated with questions related to base makeup (makeup breakdown, poor makeup application, rough texture). In this specification, this factor is called the MB factor. Factor 5 is related to the sensation of dryness and is strongly correlated with the questionnaire items related to skin dryness (dryness, preventing dryness). In this specification, this factor is referred to as Factor D, and its factor score, the degree of dryness, is denoted as SD. In this embodiment, the values ​​of each subject's two-choice answer (0 or 1) to each question were used, but the questions (and their answer values) may also be three-choice (0, 1, or 2), four-choice, five-choice, or more. Furthermore, in this embodiment, the factor scores were calculated using regression, but the method for calculating factor scores in the present invention is not limited to regression. Any calculation method that can express factor scores using a linear combination of observed quantities, such as the Bartlett method or the Anderson-Rubin method, can be used in the present invention.

[0054] (Regarding the calculation of factor scores) Here, we will provide supplementary information on how to calculate the scores for each factor of a customer and how to normalize the factor score coefficients shown in Figure 3B. The score fi for the i-th factor can generally be calculated using the following equation 4. (Equation 4) fi = Σ aij × ( xj - mj) / σj Here, aij is the (original) factor score coefficient estimated by factor analysis based on the underlying data, xj is the value of the customer's response to the j-th question, and mj and σj are the mean and standard deviation of the subject's response to the j-th question in the underlying data. Σ means the sum over j. In Equation 4, the sum over j is generally taken over all questions listed in the questionnaire related to the underlying data. However, when interviewing the customer, if some questions are selected from the questions listed in the questionnaire related to the underlying data, then in Equation 4... The sum over j is calculated over the questions selected in the medical interview. Now, when a question has only two possible answers (0 or 1), it is convenient to consider the value aij / σj (= cij), which is obtained by dividing the (original) factor score coefficient aij by the standard deviation σj. This is because, when calculating the factor score for the subject or customer, the factor score can be calculated (excluding additive constants) simply by adding cij for questions where the answer value is 1. In this specification, unless otherwise specified, cij will be called the factor score coefficient and aij will be called the (original) factor score coefficient. The same applies to Figures 9 and 11, which will be discussed later.

[0055] <1-1. A beauty assessment system that divides the skin type identification space into eight spatial regions> (3D distribution of the percentage of people who identify as having sensitive skin) The basic data includes the responses (0 or 1) to the two-choice questions asking whether the participant fits each skin type shown in Figure 2B, in addition to the 21 questions mentioned above. In particular, it includes the two-choice responses to the question asking whether the participant's skin type is sensitive (the question asking whether they perceive their skin as sensitive). Of the approximately 20,000 subjects included in the basic data, about one-third were given the question asking whether they perceive their skin as sensitive. The number of valid responses to the question asking whether they perceive their skin as sensitive was 6,069, and the respondents were randomly selected in a certain proportion for each age group in 5-year increments. Each subject included in the above basic data can be placed at a single point (coordinates (SO,SD,SH)) on a skin type identification space, where the score SO of factor O related to excess sebum is taken as the first axis (X axis), the score SD of dryness related to factor D related to dryness is taken as the second axis (Y axis), and the score SH of factor H related to the sensation of clogged pores is taken as the third axis (Z axis). It is thought that the probability of a customer who has undergone a medical interview answering YES to the question of whether they perceive themselves as having sensitive skin differs depending on their position on the skin type identification space. Therefore, the binarized score bSO of the score SO is defined as H (if it is above the average) or L (if it is below the average) depending on whether the score SO is above or below the average. Similarly, the binarized score of dryness SD that takes a value of H or L is defined as bSD, and the binarized score of the score SH is defined as bSH. Since the three binarized scores bSH, bSO, and bSD each take on either an H or L value, the skin type identification space is divided into 2 × 2 × 2 = 8 spatial regions depending on the combination of these values.

[0056] [Table 1]

[0057] Table 1 shows the percentage of subjects who answered YES to the question about whether they perceive themselves as having sensitive skin, among the subjects with valid responses included in the basic data, and who belong to each of the eight spatial domains mentioned above. Generally, the higher the SH score, the higher the SO score, and the higher the SD score for dryness, the greater the percentage of subjects who perceive themselves as having sensitive skin.

[0058] (3D distribution of the percentage of people who identify as having oily skin) Table 1 also shows the percentage of people who perceive themselves as having oily skin for each of the eight spatial domains. Generally, the higher the SH score, the higher the SO score, and the lower the SD dryness score, the higher the percentage of people who perceive themselves as having oily skin.

[0059] (3D distribution of the percentage of people who perceive themselves as having dry skin) Table 1 also shows the percentage of people who perceive themselves as having dry skin for each of the eight spatial domains. Generally, the lower the SH score, the lower the SO score, and the higher the SD dryness score, the higher the percentage of people who perceive themselves as having dry skin.

[0060] (3D distribution of the percentage of people who consider themselves to have normal skin) Table 1 also shows the percentage of people who perceive themselves as having normal skin for each of the eight spatial domains. There is a tendency for the percentage of people who perceive themselves as having normal skin to increase as the SH score, SO score, and dryness SD score decrease.

[0061] (3D distribution of the percentage of people who identify as having combination skin) Table 1 also shows the percentage of people who perceive themselves as having combination skin (dry and oily) for each of the eight spatial regions. There is a tendency for the percentage of people who perceive themselves as having combination skin to be higher as the SH score, SO score, and dryness degree SD score are higher.

[0062] (Three-dimensional distribution of the percentage of people who perceive themselves as having irregular or unhealthy eating habits) In one embodiment, as described later, the basic data includes multiple questions about dietary habits, and by performing factor analysis on the subjects' responses to these questions, the score SF for factors related to irregular eating and unhealthy eating can be calculated for each subject. If the score SF is above the average, the subject is considered to be aware of irregular eating and unhealthy eating; otherwise, they are considered not to be aware of irregular eating and unhealthy eating. Table 1 also shows the percentage of people who are aware of irregular eating and unhealthy eating for each of the eight spatial domains. There is a tendency for the percentage of people who are aware of irregular eating and unhealthy eating to be higher as the score SH is higher, the score SO is higher, and the dryness degree SD is higher. This embodiment of skin type determination means can be used as a simple screening tool to determine whether or not a customer has problems related to irregular eating and unhealthy eating.

[0063] (Three-dimensional distribution of the proportion of people who are aware of circulatory problems) In one embodiment, as described later, the basic data includes multiple questions about subjective physical condition, and by factor analysis of the subjects' responses to these questions, the score SK for factors related to circulatory dysfunction can be calculated for each subject. If the score SK is above the average value, the subject is considered to have subjective awareness of circulatory dysfunction; otherwise, they are considered not to have subjective awareness of circulatory dysfunction. Table 1 also shows the percentage of subjects who have subjective awareness of circulatory dysfunction for each of the eight spatial domains. There is a tendency for the percentage of subjects who have subjective awareness of circulatory dysfunction to be higher as the score SH is higher, the score SO is higher, and the dryness degree SD is higher. The skin type determination means of this embodiment can be used as a simple screening tool to determine whether or not a customer has problems related to circulatory dysfunction.

[0064] <1-2. Beauty assessment system using point positions in skin type identification space> (Three-dimensional distribution of the percentage of people who identify as having sensitive skin) Each subject included in the above basic data can be placed at a single point on a skin type identification space, where the score SO for factor O related to excess sebum is plotted on the first axis (X axis), the score SD for dryness related to factor D related to dryness is plotted on the second axis (Y axis), and the score SH for factor H related to pore dirt sensation is plotted on the third axis (Z axis). Based on the position on the skin type identification space, it can be determined whether the customer who underwent the interview has sensitive skin. The probability of answering YES to the awareness / awareness question is thought to be different. Figure 4-1 shows the estimation of these probabilities using the polynomial logistic regression (with regularization) method described later, and is plotted as a contour map. In the following discussion, the scores SO, SD, and SH for each factor will be normalized to values ​​between 0 and 1 by a linear transformation that converts the maximum value to 1 and the minimum value to 0. Figure 4-1A is a contour plot showing the distribution of the above probabilities in the skin type identification space on a plane where the normalized H factor score SH (hereinafter referred to as H) is 0.95. In addition to the contour lines, 100 points are plotted in this figure. These points represent 100 subjects randomly selected from the baseline data whose normalized H factor score H is between 0.90 and 1.00, with those who answered YES to the question about self-awareness of sensitive skin marked with a "○" and those who did not answer YES marked with a "+". Figure 4-1B, like Figure 4-1A, is a contour plot showing the distribution of probabilities. However, it differs from Figure 4-1A in that it is a contour plot shown on a plane where the normalized H factor score H is 0.05, and the 100 plotted points represent 100 subjects randomly selected from the baseline data whose normalized H factor score H is between 0.00 and 0.10. As can be seen from both figures, the higher the score SO for factor O related to excess sebum, and the higher the standard deviation (SD) for dryness, the greater the proportion of people who perceive themselves as having sensitive skin. Also, if the score SO for factor O related to excess sebum and the standard deviation (SD) for dryness are the same, the higher the score SH for factor H related to the sensation of clogged pores, the greater the proportion of people who perceive themselves as having sensitive skin. In other words, the higher the score SO, the higher the dryness SD, and the higher the score SH, the greater the proportion of people who perceive themselves as having sensitive skin. Furthermore, in the range where "the score H for normalized factor H is less than 0.5 and the score SO for normalized factor O is 0.8 or less" or "the score H for normalized factor H is 0.5 or higher and the score SO for normalized factor O is 0.7 or less", an increase in the standardized dryness SD tends to lead to a greater increase in the proportion of people who perceive themselves as having sensitive skin than an increase in the score SO for normalized factor O.

[0065] (3D distribution of the percentage of people who identify as having oily skin) Figures 4-2A and 4-2B are contour plots showing the distribution of the proportion of people who perceive themselves as having oily skin, estimated in the same manner. As can be seen from both figures, the proportion of people who perceive themselves as having oily skin tends to be higher as the score SO of factor O, which relates to excessive sebum, is higher, and the proportion of people who perceive themselves as having oily skin is lower. Furthermore, if the score SO of factor O, which relates to excessive sebum, and the proportion of people who perceive themselves as having oily skin are the same, the proportion of people who perceive themselves as having oily skin generally tends to be higher as the score of factor H, which relates to the sensation of clogged pores, is higher.

[0066] (3D distribution of the percentage of people who perceive themselves as having dry skin) Figures 4-3A and 4-3B are contour plots showing the distribution of the proportion of people who perceive themselves as having dry skin, estimated in the same manner. As can be seen from both figures, the proportion of people who perceive themselves as having dry skin is greatly influenced by the normalized dryness standard deviation (SD). Generally, the proportion of people who perceive themselves as having dry skin tends to be higher as the SD is larger and the score SO of factor O, which relates to excess sebum, is smaller. Also, if the score SO of factor O, which relates to excess sebum, and the SD are the same, generally, the proportion of people who perceive themselves as having dry skin tends to be higher as the score of factor H, which relates to the sensation of clogged pores, is smaller. Furthermore, when the score H of the normalized factor H is 0.5 or higher and the normalized dryness standard deviation is approximately 0.9 or higher, close to 1, there is a tendency for the proportion of people who perceive themselves as having dry skin to decrease as the SD increases, regardless of the score SO of factor O, which relates to excess sebum.

[0067] (3D distribution of the percentage of people who consider themselves to have normal skin) Figures 4-4A and 4-4B are contour plots showing the distribution of the proportion of people who perceive themselves as having normal skin, estimated in the same manner. As can be seen from both figures, the proportion of people who perceive themselves as having normal skin tends to be higher as the dryness standard deviation (SD) is smaller and the score SO of factor O, which relates to excess sebum, is smaller. Also, if the score SO of factor O, which relates to excess sebum, and the dryness standard deviation (SD) are the same, the perception of clogged pores Generally, the lower the score for factor H related to this, the higher the proportion of people who perceive themselves as having normal skin.

[0068] (3D distribution of the percentage of people who identify as having combination skin) Figures 4-5A and 4-5B are contour plots showing the distribution of the proportion of people who perceive themselves as having combination skin, estimated in the same manner. When the normalized H factor score is large, for example, 0.95, the proportion of people who perceive themselves as having combination skin reaches its maximum value when the normalized O factor score SO is approximately 0.7 and the normalized dryness SD is approximately 0.7. As SO or SD moves away from the above maximum value in terms of distance on the plane, the proportion of people who perceive themselves as having combination skin tends to decrease monotonically. When the normalized H factor score is small, for example, 0.05, the proportion of people who perceive themselves as having combination skin reaches its maximum value when the normalized O factor score SO is approximately 0.8 and the normalized dryness SD is approximately 1.0. As SO or SD moves away from the above maximum value in terms of distance on the plane, the proportion of people who perceive themselves as having combination skin tends to decrease monotonically. Furthermore, if the scores for factor O (related to excess sebum) (SO) and dryness (SD) are the same, the higher the score for factor H (related to the perception of clogged pores), the greater the proportion of people who perceive themselves as having combination skin.

[0069] (Three-dimensional distribution of the percentage of people who perceive themselves as having irregular or unhealthy eating habits) Figures 4-6A and 4-6B are contour plots showing the distribution of the proportion of individuals who perceive themselves as having irregular or unhealthy eating habits in the above embodiment, estimated in the same manner as in Figures 4-1A and 4-1B. As can be seen from both figures, the larger the standard deviation (SD) for dryness, and the larger the score (SO) for factor O related to excess sebum, the greater the proportion of individuals who perceive themselves as having irregular or unhealthy eating habits. Furthermore, if the score (SO) for factor O related to excess sebum and the standard deviation (SD) for dryness are the same, the larger the score for factor H related to the sensation of clogged pores, the greater the proportion of individuals who perceive themselves as having irregular or unhealthy eating habits. Furthermore, both an increase in the normalized O factor score SO and an increase in the normalized dryness SD contribute to an increase in the proportion of people who perceive themselves as having irregular or unhealthy eating habits. However, when the normalized H factor score H is large, for example, 0.95, the latter (increase in normalized dryness SD) contributes more, while when the normalized H factor score H is small, for example, 0.05, the former (increase in the normalized O factor score SO) contributes more. This skin type determination method can be used as a simple screening tool to determine whether a customer has problems related to irregular or unhealthy eating habits.

[0070] (Three-dimensional distribution of the percentage of people who are aware of having circulatory problems) Figures 4-7A and 4-7B are contour plots showing the distribution of the percentage of individuals experiencing circulatory problems in the above embodiment, estimated in the same manner as in Figures 4-1A and 4-1B. As can be seen from both figures, the larger the dryness SD, and the larger the score SO of factor O related to excess sebum, the greater the percentage of individuals experiencing circulatory problems. Furthermore, if the score SO of factor O related to excess sebum and the dryness SD are the same, the larger the score of factor H related to the sensation of clogged pores, the greater the percentage of individuals experiencing circulatory problems. The skin type determination means of this embodiment can be used as a simple screening tool to determine whether or not a customer has problems related to circulatory problems.

[0071] (Probability estimation using polynomial logistic regression - 3-dimensional case) This document explains how to estimate probabilities using polynomial logistic regression in a skin type identification space (which has three coordinate axes: X, Y, and Z), using the estimation of the proportion of people who perceive themselves as having sensitive skin (i.e., the probability that a customer is aware of having sensitive skin) as an example. The number of valid responses is set to m=6069, and the subjects related to the valid responses included in the basic data are represented as i=1,2,··,m. The scores of the O factor SO, dryness SD, and H factor SH of the i-th subject are normalized as Xi, Yi, and Zi, respectively, and these are collectively represented as a three-dimensional number vector xi=(Xi,Yi,Zi). Furthermore, yi is a binary data value that takes the value 1 if the i-th subject perceives themselves as having sensitive skin, and 0 otherwise. The normalized scores of each factor calculated from a customer's questionnaire data are represented as a number vector x=(X, Given a case represented by Y,Z), we want to estimate the probability p(x) that a customer is aware of having sensitive skin. Θ is the parameter to be estimated for this probability. (Equation 5) p(x) := p(Θ,x) := g(f(Θ,x)) Let us assume the form is as follows. (Note that the symbol ":=" means that the quantity on the left side is defined by the right side.) Here, f(Θ,x) is a 6th degree polynomial of three variables X, Y, Z, i.e., the number vector x=(X,Y,Z). Θ is a collective representation of its coefficients. Here, a 4th degree polynomial of three variables has 35 coefficients, including the constant term, and a 6th degree polynomial of three variables has 84 coefficients. Let t be any real number and exp be the exponential function, then g(t) is the sigmoid function (also known as the logistic function) defined by the following equation. (Formula 6) g(t) := 1 / (1+exp(-t)) The parameter Θ is determined to minimize the following cost function J = J(Θ,λ). (Equation 7) J(Θ,λ) := λ×|Θ| 2 / m + Σ[-yi×lоg(p(Θ,xi)) -(1-yi)×log(1-p(Θ,xi))] / m Here, log is the natural logarithm, Σ is the sum over i (sum over the subjects), and |Θ| 2 λ is the sum of the squares of the 83 coefficients of the 6th degree polynomial, excluding the constant term, and the positive real number λ is a regularization parameter introduced to avoid overfitting (high variance).

[0072] (Determination of regularization parameter λ and threshold c) The method for determining the regularization parameter λ is explained below. The basic data corresponding to 6069 subjects who provided valid responses is randomly divided into two sets: training data (consisting of data from 4069 subjects) and validation data (consisting of data from 2000 subjects). First, the regularization parameter λ is fixed to one of the values ​​0.01, 0.03, 0.1, 0.3, 1, 3, 10, 30, and 100, and the parameter Θ=Θ(λ) that minimizes the cost function J(Θ,λ) for the training data (m=4049) is determined. Next, using the determined parameter Θ(λ), whether each of the 2000 subjects (j) in the validation data is aware of having sensitive skin is predicted as follows. When p(Θ(λ),xj)≧c, it is predicted that the person is aware of having sensitive skin. When p(Θ(λ), xj) < c, it is predicted that the subject is not aware of having sensitive skin. Here, c is a threshold value that takes a value between 0 and 1. By fixing the threshold value c to any one of the values 0.02, 0.04, 0.06, 0.08, 0.10, 0.15, 0.20, 0.30, 0.40, 0.50, 0.60, 0.70, 0.80, and 0.90 and comparing the above prediction with the value yj representing the presence or absence of the subject's actual awareness of sensitive skin, an accuracy index representing the accuracy of this binary classification prediction, for example, the Matthews correlation coefficient, can be calculated. Once the pair of values (λ, c) of λ and c is determined, one accuracy index is determined. Therefore, while changing λ and c so that they take any of the above values, a pair of values (λ, c) of λ and c that maximizes (or minimizes) the accuracy index can be obtained. Table 2 below shows the regularization parameter λ and the threshold value c that maximize the Matthews correlation coefficient. Table 2 also includes the correct prediction rate of the entire basic data, which combines the learning data and the verification data, at the determined values of λ and c. Here, the Matthews correlation coefficient is used as the accuracy index for the binary classification prediction, but other known accuracy indexes such as the f1 score may also be used.

[0073]

Table 2

[0074] (Applicable to any binary data)[[ID=十八]] Above, the method for estimating probability was explained using the presence or absence of self-awareness of sensitive skin as an example, but this method for estimating probability can be applied to any binary data yi related to each subject (i). The binary data yi may be, for example, the presence or absence of self-awareness of oily skin, dry skin, normal skin, combination skin, irregular / unhealthy eating habits, or circulatory problems. Furthermore, the binary data yi does not have to be a quantity related to the presence or absence of self-awareness of the subject or customer. For example, it may be binary data representing the result of a judgment made by a staff member of a beauty product or service store, a doctor, a chiropractor, or other professional who looks at a questionnaire of the subject or customer, or / or directly interviews and examines them, to determine whether or not they belong to a certain skin type; binary data representing the result of a judgment made on whether or not they belong to a certain attribute related to dietary habits; or binary data representing the result of a judgment made on whether or not they belong to a certain attribute related to physical condition. In other words, the basic data of the present invention may include binary data indicating whether each subject falls under a predetermined attribute related to skin type, a predetermined attribute related to beauty, a predetermined attribute related to diet, or a predetermined attribute related to physical condition. The skin type determination means provided by the computer control program of the present invention may have means for estimating the probability that a customer falls under a predetermined attribute related to the binary data, according to the customer's position in the skin type identification space calculated from the customer's medical history data, using a polynomial logistic regression method. Note that a 4th-degree polynomial, an 8th-degree polynomial, or a 10th-degree polynomial may be used instead of a 6th-degree polynomial. Furthermore, instead of the polynomial logistic regression method, other nonlinear statistical estimation methods, such as a multilayer neural network whose output layer is composed of sigmoid neurons, or a support vector machine with sigmoid neurons added as an output layer, may be used to estimate the above probability.

[0075] In this embodiment of the beauty judgment system of the present invention, by using the scores SO, SD, and SH of three factors, the skin type of the customer can be determined more precisely according to the position of the point (SO, SD, SH) in the skin type identification space with the xyz coordinate axes. Therefore, it is possible to clarify in more detail the content of the skin type related to whether the customer has a feeling of sensitive skin, and it can be used to provide more appropriate beauty products and beauty services, as well as advice on diet and physical condition. In the examples shown in FIGS. 4-1 to 4-7, only the cases where the score H of the standardized H factor is 0.05 or 0.95 are illustrated, but the cases where the score H takes other values such as 0.50 can be illustrated in the same way. That is, in the modified form of this embodiment, when the scores of the three factors of the customer are SO, SD, and SH in order, a plane parallel to the xy plane with the z coordinate equal to the score SH in the skin type identification space with the xyz coordinate axes is used as the skin type identification section, and the skin type of the customer is determined more precisely according to the position of the point (SO, SD) in the skin type identification section. Therefore, it is possible to clarify in more detail the content of the skin type related to whether the customer has a feeling of sensitive skin, and it can be used to provide more appropriate beauty products and beauty services, as well as advice on diet and physical condition. On the skin type identification section, in addition to the mark indicating the point (SO, S D) related to the customer, it is preferable to display a distribution graph such as a contour line or a heat map showing the probability distribution of the customer being a specific skin type.

[0076] <2. Beauty Judgment System for Determining Skin Type by Two Factors> Next, the beauty judgment for determining the skin type by two factors will be described. Using the score SO of the O factor related to sebum excess, the score SH of the H factor related to the feeling of clogged pores, and an angle t such that 0° < t < 90°, the composite lipidicity SC can be calculated by Equation 1. The composite lipidicity SC is a quantity that aggregates the two factor scores SО and SH.

[0077] <2-1. Beauty Judgment System for Dividing the Skin Type Identification Plane into 10×10 Grid Squares> Figure 5A is a heatmap showing the percentage of people who answered that their skin type is "sensitive skin" for each of the 10x10 squares on the skin type identification plane, when the combined oiliness SC and dryness SD are each represented on a scale of 1 to 10 by dividing the range between the maximum and minimum values ​​into 10 equal parts using the method described earlier in "Calculation of Correlation Coefficient R". Figure 5B is a heatmap showing the percentage of people who answered that their skin type is "oily skin", Figure 5C is "dry skin", Figure 5D is "normal skin", and Figure 5E is "combination skin (dry and oily)". Figure 5F is a heatmap showing the number of subjects belonging to each square. Note that in Figures 5A to 5F, calculations were performed with an angle t of 45°. Figure 5A shows that the higher the combined oiliness score (SC) and the higher the dryness score (SD), the higher the percentage of people who perceive themselves as having "sensitive skin." Figure 5B shows that the higher the combined oiliness score (SC), the higher the percentage of people who perceive themselves as having "oily skin." Also, the lower the dryness score (SD), the slightly higher the percentage of people who perceive themselves as having "oily skin." Figure 5C shows that the higher the dryness score (SD) and the lower the combined oiliness score (SC), the higher the percentage of people who perceive themselves as having "dry skin." Figure 5D shows that the lower the dryness score (SD) and the lower the combined oiliness score (SC), the higher the percentage of people who perceive themselves as having "normal skin." Figure 5E shows that the higher the combined oiliness score (SC), the higher the percentage of people who perceive themselves as having "dry and oily combination skin." Also, the higher the dryness score (SD), the slightly higher the percentage of people who perceive themselves as having "dry and oily combination skin." These figures show that the dryness degree (SD) and the aggregated compound oiliness degree (SC) statistically explain the subject's self-perceived skin type well. In this embodiment, the position of a point (SC,SD) on a skin type identification plane equipped with XY coordinate axes is represented by one of the squares arranged in a 10x10 grid, visually representing the customer's skin type. This allows both the customer and the seller to quickly grasp and share information about the customer's skin type. Furthermore, while referring to the customer's self-perceived skin type, the customer and / or the seller can understand the customer's skin type more precisely and accurately based on the position of the point (SC,SD) on the skin type identification plane, and use this information to help them purchase and recommend beauty products.

[0078] <2-2. A beauty assessment system that estimates the probability distribution on the skin type identification plane> (Two-dimensional distribution of the percentage of people who identify as having sensitive skin) As previously described, the basic data includes the two-choice answers each subject gave to the question regarding whether they perceive themselves as having sensitive skin. Each subject included in the basic data can be placed at a single point on a skin type identification plane, with combined oiliness SC as the first axis (X-axis) and dryness SD as the second axis (Y-axis). It is thought that the probability of a customer answering YES to the question regarding whether they perceive themselves as having sensitive skin differs depending on their position on the skin type identification plane. This probability was estimated using the polynomial logistic regression (with regularization) method described later and is shown as a contour plot in Figure 6-1. In the following, both combined oiliness SC and dryness SD are normalized to a range of 0 to 1 by a linear transformation that converts the maximum value to 1 and the minimum value to 0 for consideration. In addition to the contour lines, 100 points are plotted in Figure 6-1. These points were determined by randomly selecting 100 subjects from the baseline data and marking those who answered YES to the question about whether they perceived themselves as having sensitive skin with a "○" mark. Those who do not fit this description are plotted with a "+" mark. In addition, in Figures 6-1 to 6-7, the angle t in the definition formula (Equation 1) for combined oiliness SC was set to 35°. This angle t is the angle that maximizes the correlation coefficient shown in Figure 7 related to the perception of sensitive skin, as will be explained later. As can be seen from Figure 6-1, individuals who perceive themselves as having sensitive skin are concentrated in the upper zone of the figure. The proportion of individuals who perceive themselves as having sensitive skin tends to increase as the combined oiliness score (SC) and the dryness score (SD) increase. Furthermore, in the region where the normalized dryness score (SD) is 0.5 or higher, or where the normalized combined oiliness score (Normalized Combined Oiliness) is 0.8 or lower, an increase in normalized dryness score (SD) tends to lead to a greater increase in the proportion of individuals who perceive themselves as having sensitive skin than an increase in normalized combined oiliness score (SC). Sensitive skin is basically triggered by a dry tendency, but it can also be triggered by a dry and oily tendency. The former has been known for some time, but the latter is a new finding.

[0079] (Two-dimensional distribution of the percentage of people who identify as having oily skin) Figure 6-2 is a contour plot showing the distribution of the proportion of people who perceive themselves as having oily skin, estimated in the same manner. People who perceive themselves as having oily skin are concentrated in the lower right zone of the figure. In the region where the normalized combined oiliness score SC is less than 0.8, the proportion of people who perceive themselves as having oily skin tends to increase as the combined oiliness score and dryness score decrease. Also, in the region where the normalized combined oiliness score SC is 0.8 or higher, the proportion of people who perceive themselves as having oily skin tends to increase as the dryness score SD decreases.

[0080] (Two-dimensional distribution of the percentage of people who perceive themselves as having dry skin) Figure 6-3 is a contour plot showing the distribution of the proportion of people who perceive themselves as having dry skin, estimated in the same manner. People who perceive themselves as having dry skin are concentrated in the upper left zone of the figure. The proportion of people who perceive themselves as having dry skin is greatly influenced by the normalized dryness score (SD). Generally, the proportion of people who perceive themselves as having dry skin tends to be higher as the dryness score (SD) is larger and the combined oiliness score (SC) is smaller. Furthermore, when the normalized dryness score (SD) is close to 1 (approximately 0.8 or higher), there is a tendency for the proportion of people who perceive themselves as having dry skin to decrease as the combined oiliness score (SC) increases.

[0081] (Two-dimensional distribution of the percentage of people who consider themselves to have normal skin) Figure 6-4 is a contour plot showing the distribution of the proportion of people who perceive themselves as having normal skin, estimated in the same manner. People who perceive themselves as having normal skin are concentrated in the lower left zone of the figure. Generally, the proportion of people who perceive themselves as having normal skin tends to be higher as the dryness standard (SD) and the combined oiliness standard (SC) decrease.

[0082] (Two-dimensional distribution of the percentage of people who identify as having combination skin) Figure 6-5 is a contour plot showing the distribution of the proportion of individuals who perceive themselves as having combination skin, estimated in a similar manner. The proportion of individuals who perceive themselves as having combination skin is highest when the normalized combined oiliness (SC) is approximately 0.75 and the normalized dryness (SD) is approximately 0.70, and as SC or SD moves away from the above maximum value in terms of distance on the plane, the proportion of individuals who perceive themselves as having combination skin tends to decrease monotonically.

[0083] (Two-dimensional distribution of the percentage of people who perceive themselves as having irregular or unhealthy eating habits) In one embodiment, as will be described later, the basic data includes multiple questions about dietary habits, and by performing factor analysis on the subjects' responses to these questions, the score SF for factors related to irregular eating and unhealthy eating can be calculated for each subject. If the score SF is above the mean, the subject is considered to be aware of irregular eating and unhealthy eating; otherwise, they are considered not to be aware of irregular eating and unhealthy eating. Figure 6-6 is a contour plot showing the distribution of the proportion of people who are aware of irregular eating and unhealthy eating, estimated in the same way as Figure 6-1. The proportion of people who are aware of irregular eating and unhealthy eating tends to be higher as the combined fat content SC and the dryness SD increase. Furthermore, both an increase in the normalized combined fat content SC and an increase in the normalized dryness SD contribute to an increase in the proportion of people who are aware of irregular eating and unhealthy eating, but the former (increase in normalized combined fat content SC) contributes more than the latter (increase in normalized dryness SD). This also contributes significantly. This form of skin type determination means can be used as a simple screening tool to determine whether a customer has problems related to irregular eating or poor diet.

[0084] (Two-dimensional distribution of the proportion of people who are aware of having circulatory problems) In one embodiment, as will be described later, the basic data includes multiple questions about subjective perception of physical condition, and by performing factor analysis on the subjects' responses to these questions, the score SK for factors related to circulatory dysfunction can be calculated for each subject. If the score SK is above the average value, the subject is considered to have subjective perception of circulatory dysfunction; otherwise, they are considered not to have subjective perception of circulatory dysfunction. Figure 6-7 is a contour plot showing the distribution of the proportion of persons who have subjective perception of circulatory dysfunction, estimated in the same manner as in Figure 6-1. The proportion of persons who have subjective perception of circulatory dysfunction tends to be higher as the combined oiliness SC and the dryness SD are higher. Both the increase in the normalized combined oiliness SC and the increase in the normalized dryness SD contribute to the increase in the proportion of persons who have subjective perceptions of irregular eating and poor eating habits, and the degree of contribution of the two is about the same. The skin type determination means of this embodiment can be used as a simple screening tool to determine whether or not a customer has problems related to circulatory dysfunction.

[0085] (Probability estimation using polynomial logistic regression - 2-dimensional case) This document explains how to estimate probabilities using polynomial logistic regression on a skin type identification plane (which has two coordinate axes, X and Y), using the estimation of the proportion of people who perceive themselves as having sensitive skin (i.e., the probability that a customer is aware of having sensitive skin) as an example. The number of valid responses is set to m=6069, and the subjects related to the valid responses included in the basic data are represented as i=1,2,··,m. The combined oiliness SC and dryness SD of the i-th subject are normalized as described above and denoted as Xi and Yi, respectively, and these are collectively represented as a two-dimensional number vector xi=(Xi,Yi). Furthermore, yi is a binary data value that takes the value 1 if the i-th subject perceives themselves as having sensitive skin, and 0 otherwise. Given that the normalized scores of each factor calculated from a customer's questionnaire data are represented as a number vector x=(X,Y) on the skin type identification plane, we want to estimate the probability p(x) that a customer perceives themselves as having sensitive skin. Θ is the parameter to be estimated, and the probability we are seeking is... (Equation 8) p(x) := p(Θ,x) := g(f(Θ,x)) Assume it is in the form of... Here, f(Θ,x) is a sixth-degree polynomial in two variables X and Y, i.e., the number vector x=(X,Y). Θ represents its coefficients collectively. Here, a fourth-degree polynomial in two variables has 15 coefficients, including the constant term, while a sixth-degree polynomial in two variables has 28 coefficients. g is the sigmoid function (also known as the logistic function). The parameter Θ is determined to minimize the following cost function J = J(Θ,λ). (Equation 9) J(Θ,λ) := λ×|Θ| 2 / m + Σ[-yi×lоg(p(Θ,xi)) -(1-yi)×log(1-p(Θ,xi))] / m Here, log is the natural logarithm, Σ is the sum over i (sum over the subjects), and |Θ| 2 λ is the sum of the squares of the 27 coefficients of the 6th degree polynomial, excluding the constant term, and the positive real number λ is a regularization parameter introduced to avoid overfitting (high variance).

[0086] (Determination of regularization parameter λ and threshold c) The method for determining the regularization parameter λ and threshold c is the same as in the case of the 3D skin type discrimination space, so a detailed explanation is omitted. Table 3 shows the regularization parameter λ and threshold c that maximize the Matthews correlation coefficient. Table 3 also includes the prediction accuracy on the entire base data, combining the training data and validation data, for the determined values ​​of λ and c. Here, the Matthews correlation coefficient was used as an accuracy metric for binary classification prediction, but other known accuracy metrics such as the f1 score may also be used.

[0087] [Table 3]

[0088] As can be seen by comparing Table 3 and Table 2, the two-dimensional analysis on the skin type identification plane using two variables, including the aggregated composite oiliness score SC (Table 3), is able to estimate the probability distribution for each binary data with an accuracy rate almost comparable to the three-dimensional analysis on the skin type identification space using three variables (Table 2). However, there is a slight decrease in accuracy for the presence or absence of self-perception of oily skin and the presence or absence of self-perception of normal skin. This means that in order to accurately identify self-perception of skin type related to excessive sebum and pore dissatisfaction, it is preferable to perform the analysis using three variables rather than two.

[0089] (The analysis on the skin type identification plane can be applied to any binary data.) Above, using the presence or absence of self-awareness of sensitive skin as an example, a method for estimating probabilities on the skin type identification plane using polynomial logistic regression was explained. This probability estimation method can be applied to any binary data yi relating to each subject (i), similar to the probability estimation method in the skin type identification space described above. That is, the basic data of the present invention can include binary data representing whether each subject falls under a predetermined attribute related to skin type, a predetermined attribute related to beauty, a predetermined attribute related to diet, or a predetermined attribute related to physical condition. The skin type determination means provided by the computer control program of the present invention can have means for estimating the probability that the customer falls under a predetermined attribute relating to the binary data, according to the customer's position on the skin type identification plane calculated from the customer's medical history data, using the polynomial logistic regression method. Note that instead of a 6th-degree polynomial, a 4th-degree polynomial, an 8th-degree polynomial, or a 10th-degree polynomial may be used. Alternatively, instead of polynomial logistic regression, other nonlinear statistical estimation methods can be used to estimate the above probabilities, such as multilayer neural networks with output layers composed of sigmoid neurons, or support vector machines with added sigmoid neurons as output layers.

[0090] In this embodiment of the beauty determination system of the present invention, by using the composite fatness degree SC and the dryness degree SD, the skin type of the customer can be determined more precisely and in finer detail based on the position of the point (SC, SD) in the skin type discrimination plane having the XY coordinate axes. Thus, it is possible to clarify in more detail the content of the skin type related to the presence or absence of the customer's awareness of sensitive skin, and to utilize it for providing more appropriate beauty products and beauty services, as well as providing advice regarding diet and physical condition.

[0091] <3. Beauty determination system using skin quality score SQ> (Correlation coefficient R calculated from the distribution of the skin quality score SQ and the answers to the question about the presence or absence of awareness of sensitive skin) Using the composite fatness degree SC, the dryness degree SD, and an angle s such that 0° < s < 90°, the skin quality score SQ can be calculated by Equation 2. The skin quality score SQ is a quantity representing the aggregation of two quantities SC and SD. Here, as shown in Equation 1, the composite fatness degree SC depends on an angle t such that 0° < t < 90°. Therefore, the skin quality score SQ depends not only on the value of the answer of the subject or customer to the question, but also on the two parameters, the angles s and t. Figure 15A shows the results of calculating the correlation coefficient R from the distribution of skin quality scores (SQ) and responses to the question about whether or not a subject perceives sensitive skin, after discretizing the subjects' skin quality scores (SQ) in the basic data into 10 levels from 1 to 10 using the method described above. Both angles t and s were set to 45°. The line graph shows the proportion of subjects who perceive themselves as having sensitive skin among those whose skin quality scores (SQ) belong to each interval, the bar graph shows the number of subjects who perceive themselves as having sensitive skin among those whose skin quality scores (SQ) belong to each of the 10 intervals, and the straight line shows the regression line obtained using the method described above. The correlation coefficient R was found to be 0.982 (i.e., the square root of 0.9643). It can be seen that the larger the skin quality score (SQ), the larger the proportion of subjects who perceive themselves as having sensitive skin, indicating a strong correlation between skin quality scores (SQ) and whether or not a subject perceives themselves as having sensitive skin. According to the regression line, the above percentage is approximately 0% in the interval SQ01, where the skin quality score SQ is smallest, while it is approximately 59% in the interval SQ10, where the skin quality score SQ is largest. In other words, to put it somewhat dramatically, a high skin quality score SQ for a customer is almost synonymous with the customer considering themselves to have sensitive skin.

[0092] (Relationship between angles t,s and correlation coefficient R) Figure 7 is a table showing the correlation coefficient R calculated from the distribution of responses to the question about whether or not a person perceives their skin as sensitive, when the angles t and s, which are two parameters appearing in the definition formulas of the skin quality score SQ and the combined oiliness score SC, are set to various angles, including angles other than 45°. The correlation coefficient R takes its maximum value of R = 0.996 at (t,s) = (35°,40°). If we set a correlation coefficient R of 0.94 or higher as one condition, then this condition is met if 25°≦t≦50° and 25°≦s≦75°. If we set a correlation coefficient R of 0.98 or higher as another condition, then this condition is met if 20°≦t≦40° and 40°≦s≦60°. When the angles t and s are set so that the above correlation coefficient R is close to 1, it can be said that the skin quality score SQ is a quantity that has an extremely strong correlation with whether or not a person perceives their skin as sensitive.

[0093] <4. The relationship between diet and self-awareness of sensitive skin> (Factor analysis of basic dietary data) Of the subjects included in the aforementioned basic data, 6069 who were asked about their self-awareness of having sensitive skin were further questioned about their dietary habits, and their responses (0 or 1) were recorded. The questions consisted of 25 multiple-choice questions, n01 to n25, as shown in the table in Figure 8. The answers to these questions constitute the basic dietary data. Similar to the skin diagnosis, eight factors were extracted using factor analysis of an orthogonal model. Factor 1 can be interpreted as health consciousness, Factor 2 (hereinafter referred to as the F2 factor) as unhealthy (tendency to overeat), Factor 3 (hereinafter referred to as the F1 factor) as irregular eating habits, Factor 4 as prevention of lifestyle-related diseases, Factor 5 as supplement preference, Factor 6 as preference for convenience, Factor 7 as low processing (minimal cooking), and Factor 8 as a factor related to menu (three meals / gourmet food, little dessert / little snacks). Factor weighting coefficients are omitted. The factor score coefficients obtained by regression are shown in the table in Figure 9.

[0094] (The relationship between irregular eating habits and self-awareness of sensitive skin) Figure 15B shows the results of calculating the correlation coefficient R from the distribution of factor scores SF1 and responses to the question about whether or not a subject perceives sensitive skin, using the method described above to discretize the F1 factor score SF1 related to irregular eating habits of the subjects in the basic data into 10 levels from 1 to 10. The line graph shows the proportion of subjects who perceive themselves as having sensitive skin among those whose factor score SF1 belongs to each interval, the bar graph shows the number of subjects who perceive themselves as having sensitive skin among those whose factor score SF1 belongs to each of the 10 intervals, and the straight line shows the regression line obtained using the method described above. The correlation coefficient R was found to be 0.847 (i.e., the square root of 0.7208). It can be seen that the F1 factor score SF1 related to irregular eating habits has a moderately strong correlation with whether or not a subject perceives themselves as having sensitive skin. The larger the factor score SF1, the larger the proportion of subjects who perceive themselves as having sensitive skin. According to the regression line, in the interval SF1_01 where the factor score SF1 is smallest, the above proportion is approximately 19%. In contrast, in the interval SF1_10, where the factor score SF1 is the largest, the above proportion is approximately 36%.

[0095] (The relationship between an unhealthy diet and self-awareness of sensitive skin) Fig. 15C is a diagram showing the calculation of the correlation coefficient R calculated from the distribution of the factor score SF2 of the F2 factor related to the subject's no-healthy (overeating tendency), that is, an unhealthy diet in the basic data discretized into 10 levels from 1 to 10 by the above method and the answers to the question on whether or not the subject has a sensitive skin awareness. The line graph shows the percentage of subjects who are aware of having sensitive skin among the subjects whose factor score SF2 belongs to each interval, the bar graph shows the number of subjects who are aware of having sensitive skin among the subjects whose factor score SF2 belongs to each of the 10 intervals, and the straight line shows the regression line obtained by the above method. The correlation coefficient R was determined to be 0.856 (that is, the square root of 0.7327). It can be seen that the factor score SF2 related to an irregular diet has a somewhat strong correlation with the presence or absence of sensitive skin awareness. The larger the factor score SF2, the larger the percentage of subjects who are aware of having sensitive skin. According to the regression line, in the interval SF2_01 where the factor score SF2 is the smallest, the above percentage is about 18%, while in the interval SF2_10 where the factor score SF2 is the largest, the above percentage is about 33%.

[0096] (Relationship between the score SF of the composite diet factor and the awareness of sensitive skin) Using the factor score SF1, the factor score SF2, and an angle u such that 0° < u < 90°, the score SF of the composite diet factor can be calculated by Equation 3. The score SF is a quantity representing the aggregation of the two quantities SF1 and SF2. As can be seen from Equation 3, the score SF depends not only on the values of the answers of the subjects or customers to the questions but also on the parameter angle u. Figure 15D shows the results of calculating the correlation coefficient R from the distribution of scores SF and responses to the question about whether or not a subject is aware of having sensitive skin, after discretizing the scores SF of the subjects in the basic data into 10 levels from 1 to 10 using the method described above. The angle u was set to 45°. The line graph shows the proportion of subjects who are aware of having sensitive skin among those whose scores SF belong to each interval, the bar graph shows the number of subjects who are aware of having sensitive skin among those whose scores SF belong to each of the 10 intervals, and the straight line shows the regression line obtained using the method described above. The correlation coefficient R was found to be 0.973 (i.e., the square root of 0.9467). It can be seen that the larger the score SF, the larger the proportion of subjects who are aware of having sensitive skin, indicating a strong correlation between the score SF and whether or not a subject is aware of having sensitive skin. According to the regression line, the above proportion is approximately 19% in the interval SF01, which has the smallest score SF, while it is approximately 40% in the interval SF10, which has the largest score SF.

[0097] (Relationship between angle u and correlation coefficient R) Table 4 below shows the correlation coefficient R calculated from the distribution of the score SF and the responses to the question about whether or not one perceives sensitive skin, when the angle u, a parameter appearing in Equation 3 which defines the score SF of the complex diet factor, is set to various angles in 5° increments, including angles other than 45°.

[0098] [Table 4]

[0099] The correlation coefficient R takes its maximum value of R = 0.988 at u = 35°. The correlation coefficient R is 0. If we set a condition that the correlation coefficient R is 96 or higher, then the condition is met if u = 5°, 15°, 20°, 30°, 35°, 40°, 45°, 55°, or 60°. If we set another condition that the correlation coefficient R is 0.98 or higher, then the condition is met if u = 20°, 30°, 35°, or 40°. If the angle u is not divisible by 5°, for example, we can determine whether the angle u satisfies the condition by linearly interpolating the values ​​of the correlation coefficient R at 5° intervals in the table above. When the angle u is set so that the correlation coefficient R takes a value close to 1, it can be said that the score SF of the composite diet factor is a quantity that has a strong correlation with the presence or absence of self-awareness of sensitive skin.

[0100] <5. The relationship between physical condition and self-perception of sensitive skin> (Factor analysis of basic health data) Of the subjects included in the aforementioned basic data, 6069 individuals who were asked about their self-awareness of sensitive skin were further questioned about their self-awareness of their physical condition, and their responses (0 or 1) were recorded. The questions consisted of 30 multiple-choice questions, h01 to h25, as shown in the table in Figure 10. The answers to these questions constitute the basic health data. Similar to the skin diagnosis, 14 factors were extracted using factor analysis with an orthogonal model. Factor 1 (hereinafter referred to as the MEM factor) can be interpreted as a factor related to mental state. Factor 2 (hereinafter referred to as the KT factor) can be interpreted as a factor related to high activity, Factor 3 (hereinafter referred to as the CH factor) as good or enhanced condition, Factor 4 (hereinafter referred to as the EYE factor) as eye strain, Factor 5 (hereinafter referred to as the A factor) as menopausal age, Factor 6 (hereinafter referred to as the T factor) as menopausal specificity, Factor 7 (hereinafter referred to as the GE factor) as unclear or ambiguous cause, Factor 8 (hereinafter referred to as the KK factor) as discontinuous or repetitive period, Factor 9 (hereinafter referred to as the SM factor) as sleep, Factor 10 (hereinafter referred to as the HR factor) as fatigue, Factor 11 (hereinafter referred to as the AL factor) as allergies, Factor 12 (hereinafter referred to as the IT factor) as pain, Factor 13 (hereinafter referred to as the NE factor) as mucous membranes, and Factor 14 (hereinafter referred to as the KE factor) as circulatory disorders of blood, lymph, etc. The estimated factor weighting coefficients are omitted. The factor score coefficients obtained by the regression method are shown in the table in Figure 11. In the factor analysis described above, two factors, Factor A and Factor T, were extracted, which are related to symptoms common in menopause. Factor A can be interpreted as a factor corresponding to menopausal physical conditions that appear in relation to the subject's age, while Factor T can be interpreted as a factor corresponding to menopausal physical conditions that appear regardless of the subject's age. So-called premature menopausal symptoms are mainly related to Factor T.

[0101] (Scores related to menopausal age factors and self-awareness of sensitive skin) Figure 15E shows the results of calculating the correlation coefficient R from the distribution of factor scores SA and responses to the question about whether or not a subject is aware of having sensitive skin, using the method described above to discretize the factor scores SA of the subjects in the basic data into 10 levels from 1 to 10. The line graph shows the proportion of subjects who are aware of having sensitive skin among those whose factor scores SA belong to each interval, the bar graph shows the number of subjects who are aware of having sensitive skin among those whose factor scores SA belong to each of the 10 intervals, and the straight line shows the regression line obtained using the method described above. The correlation coefficient R was found to be -0.511 (i.e., the square root of 0.2613). It can be seen that the factor score SA of factor A (menopausal age factor) has some correlation with whether or not a subject is aware of having sensitive skin. The larger the factor score SA, the smaller the proportion of subjects who are aware of having sensitive skin. According to the regression line, the above proportion is approximately 27% in the interval SA01, where the factor score SA is smallest, while it is approximately 21% in the interval SA10, where the factor score SA is largest.

[0102] (Scores on menopausal-specific factors and self-awareness of sensitive skin) Figure 15F shows the results of calculating the correlation coefficient R, which is calculated from the distribution of factor scores ST and responses to the question about whether or not the subject perceives sensitive skin, using the method described above to discretize the factor scores ST for the subjects in the basic data related to the T factor (a menopausal physical condition factor that appears regardless of the subject's age). The line graph shows the proportion of subjects whose factor scores ST belong to each interval. The graph shows the percentage of subjects who are aware of having sensitive skin. The bar graph shows the number of subjects who are aware of having sensitive skin among those belonging to each of the 10 intervals for factor score ST. The straight line shows the regression line obtained using the method described above. The correlation coefficient R was found to be 0.886 (i.e., the square root of 0.7848). It can be seen that the score ST of factor T has a moderately strong correlation with the presence or absence of self-awareness of sensitive skin. The larger the factor score ST, the larger the percentage of subjects who are aware of having sensitive skin. According to the regression line, the above percentage is approximately 17% in the interval ST01, which has the smallest factor score ST, while it is approximately 36% in the interval ST10, which has the largest factor score ST.

[0103] (Circulation factor score and self-awareness of sensitive skin) Figure 15G shows the results of calculating the correlation coefficient R from the distribution of factor scores SK and responses to the question about whether or not a subject is aware of having sensitive skin, using the method described above to discretize the factor scores SK of the subjects in the basic data for circulatory factors (factors related to circulatory disorders such as blood and lymph). The line graph shows the proportion of subjects who are aware of having sensitive skin among those whose factor scores SK belong to each interval, the bar graph shows the number of subjects who are aware of having sensitive skin among those whose factor scores SK belong to each of the 10 intervals, and the straight line shows the regression line obtained using the method described above. The correlation coefficient R was found to be 0.967 (i.e., the square root of 0.9359). It can be seen that the factor score SK has a strong correlation with whether or not a subject is aware of having sensitive skin. The larger the factor score SK, the larger the proportion of subjects who are aware of having sensitive skin. According to the regression line, the above proportion is approximately 15% in the interval SK01 where the factor score SK is smallest, while it is approximately 47% in the interval SK10 where the factor score ST is largest.

[0104] (I am aware of my allergies and sensitive skin.) Figure 15H shows the results of calculating the correlation coefficient R from the distribution of factor scores SAL related to the allergy-related AL factor for the subjects in the basic data, discretized into 10 levels from 1 to 10 using the method described above, and the correlation coefficient R calculated from the distribution of factor scores SAL and responses to the question of whether or not the subject perceives sensitive skin. The line graph shows the proportion of subjects who perceive sensitive skin among those whose factor scores SAL belong to each interval, the bar graph shows the number of subjects who perceive sensitive skin among those whose factor scores SAL belong to each of the 10 intervals, and the straight line shows the regression line obtained using the method described above. The correlation coefficient R was found to be 0.858 (i.e., the square root of 0.7357). It can be seen that the factor score SAL has a moderately strong correlation with whether or not the subject perceives sensitive skin. The larger the factor score SAL, the larger the proportion of subjects who perceive sensitive skin. According to the regression line, the above proportion is approximately 13% in the interval SAL01, where the factor score SAL is smallest, while it is approximately 50% in the interval SAL10, where the factor score ST is largest.

[0105] <6. Relationship between skin quality score (SQ) and diet or self-perceived physical condition> As shown in Equation 2, the skin quality score SQ is an aggregated value obtained by multiplying the combined oiliness and dryness by positive weights and adding them together. Figures 6-6 and 6-7 suggest a strong correlation between the skin quality score SQ and the various factor scores related to diet or perceived physical condition. Table 5 below shows the statistical relationship between the skin quality score SQ and the various factor scores related to diet or perceived physical condition. Here, the angles t and s in Equations 1 and 2 were calculated assuming t=s=45°.

[0106] [Table 5]

[0107] First, let's explain the first theme in Table 5, "Skin Type Score and Irregular Eating Habits." Depending on whether the subject's score for irregular eating habits (SF1) included in the basic data is above the mean, SF1 is binarized to either 1 or 0 (binary data), and this is represented as bSF1. By using the binarized score for irregular eating habits (bSF1) instead of the answer to the question about whether or not the subject perceives sensitive skin, a graph similar to Figure 15A can be drawn, and the correlation coefficient can be calculated. The correlation coefficient R calculated from the distribution of skin type score SQ and binary data bSF1 is 0.961, which is a large value close to 1. Furthermore, when considering the regression line, among the 10 intervals obtained by dividing the range between the maximum and minimum values ​​of the skin quality score SQ into 10 parts, the proportion of subjects whose binary data bSF1 is 1 in the first interval, where the skin quality score SQ is smallest, is 20.4%, while the proportion of subjects whose binary data bSF1 is 1 in the 10th interval, where the skin quality score SQ is largest, is 54.9%, which is approximately 2.5 times higher than in the first interval. In this sense, there is a strong correlation between the skin quality score SQ and the binarized dietary irregularity factor score bSF1.

[0108] The same applies to the remaining themes in Table 5, where there are strong statistical correlations between the skin quality score SQ and the binarized score bSF2 of unhealthy eating factors, the skin quality score SQ and the binarized score bSF of combined eating factors (here, angle u is calculated as 45° as an example), the skin quality score SQ and the binarized score bST of menopausal specific factors, the skin quality score SQ and the binarized score bSK of circulatory factors, and the skin quality score SQ and the binarized score bSAL of allergy factors. However, the statistical correlation between the skin quality score SQ and the binarized score bSA of menopausal age factors is small. The beauty judgment program in the beauty judgment system of the present invention may be equipped with means for outputting advice regarding the customer's diet or physical condition to an output device based on which of the above 10 intervals the customer's skin quality score SQ, calculated from the questionnaire data, belongs to, and the proportion of subjects in that interval where various binary data take a value of 1, or the value of that proportion estimated by the above regression line.

[0109] <7. Configuration of the Beauty Assessment System> <7-1. Independent Systems> Figures 12A and 12B are configuration diagrams showing the system configuration in one embodiment of the present invention, which is composed of an independent computer. The beauty judgment system of this embodiment comprises an independent computer 10 such as a personal computer, an input device 20 consisting of a keyboard, mouse, touch panel display, etc., an output device 40 consisting of a display, printer, etc., an internal storage means 31 such as CPU registers, RAM, ROM, etc., and / or a computer-readable storage means 30 consisting of a hard disk drive (HDD) or solid state drive (SSD) or removable storage device such as a USB memory or SD card or an external storage means (storage medium) 32 such as a CD-R or DVD. The system consists of the following components. This beauty assessment system may be operated by the customer, or by a beauty advisor at a beauty supply store, etc., while interviewing the customer. The input device 20 is used to input each customer's interview data. Here, each customer's interview data may include answers to questions about beauty, whether or not they use beauty products, answers to questions about their diet, answers to questions about their physical condition, age, and other information. The computer 10 executes a computer control program 50 stored in the storage means 30 to perform a beauty assessment of the customer based on the analyzed parameters 33 stored in the storage means 30 and the interview data, and outputs the assessment result to the output device 40. The analyzed parameters 33 are obtained in advance by analyzing the basic data using a factor analysis method, and this analysis may be performed by the computer 10 or by another computer. In the latter case, the analyzed parameters 33 obtained by the analysis by the other computer can be read by the computer 10 via an external storage means (storage medium) 32 or via a network. The above assessment results may include not only the results of beauty assessments such as skin type, but also recommendations for beauty products, dietary advice, and health-related advice.

[0110] (Basic computer configuration) Figure 12B shows a preferred basic configuration 11 (basic configuration A) for the computer 10. Basic configuration A consists of an input device 20 for inputting customer medical history data, a computer-readable storage means 30 for storing basic data, analyzed parameters 33, and a computer control program 50, and for temporarily storing medical history data, and an output device 40 for outputting judgment results such as beauty judgments. The computer 10 performs calculations of factor scores related to beauty, judgments such as beauty judgments, and output of judgment results by executing the computer control program 50. The storage means 30 consists of an internal storage means 31 and / or an external storage means 32. The internal storage means 31 is, for example, a CPU register, RAM, ROM or other memory. The external storage means (storage medium) 32 is, for example, a hard disk drive (HDD), a solid state drive (SSD), a removable storage device such as a USB memory or SD card, or a CD-R or DVD. The analyzed parameters 33 refer to parameters related to the computer control program, and include information other than raw questionnaire data and raw basic data that is necessary for performing cosmetic assessments. These include standardization parameters such as the mean and standard deviation of the answer values ​​to questions in the basic data, factor weights and factor score coefficients found through factor analysis, factor scores for each quantile, thresholds used for classification and recommendation discrimination, coefficients and regularization parameters of polynomials used for estimation, and connection weights in neural networks.

[0111] (Computer-controlled program) Figure 13 is a configuration diagram showing the configuration of a computer control program 50 in one embodiment of the present invention. The computer control program 50 includes a beauty judgment program 60, a dietary lifestyle analysis program 70, a physical condition analysis program 80, and a result output program 90. The beauty judgment program 60 has a skin type judgment means (subprogram) 61 for performing a beauty judgment on a customer based on analyzed parameters 33 stored in a storage means 30 and the customer's medical history data. The dietary lifestyle analysis program 70 has a dietary lifestyle judgment means (subprogram) 71 for performing a dietary lifestyle judgment on a customer based on analyzed parameters 33 related to dietary lifestyle stored in a storage means 30 and the customer's medical history data. The physical condition analysis program 80 has a physical condition judgment means (subprogram) 81 for performing a physical condition judgment on a customer based on analyzed parameters 33 related to the customer's perceived physical condition stored in a storage means 30 and the customer's medical history data. The result output program 90 has a judgment result output means 91 for outputting the results of the customer's skin type judgment, dietary lifestyle judgment, and / or physical condition judgment to an output device 40.

[0112] <7-2. System for performing beauty assessments and other evaluations on a server> Figure 14A is a diagram showing the system configuration in another embodiment of the present invention. In this embodiment, as a general rule, judgments such as beauty assessments are performed by the server 10a, and input and output are performed by the terminal 10b, which is closer to the customer. The beauty assessment system according to this embodiment includes a server 10a and a terminal 10b that can communicate bidirectionally via a wired or wireless network 12 such as an intranet or the internet. The terminal 10b is a computer such as a personal computer, smartphone, or tablet. The terminal 10b is connected to an input device 20b consisting of a keyboard, mouse, touch panel display, etc., and an output device 40b consisting of a display, printer, etc. The server 10a is a computer such as a personal computer or server machine, and has the basic configuration A described above.

[0113] Terminal 10b may be operated by the customer, or by a beauty advisor such as a beauty supply store while interviewing the customer. Terminal 10b stores in its memory means a medical interview data input means (subprogram) 51 for inputting customer medical interview data using the input device 20b, and a medical interview data transmission means (subprogram) 52 for transmitting the input medical interview data to the server 10a via the network 12. Server 10a stores in its memory means a medical interview data receiving means (subprogram) 54 for receiving customer medical interview data transmitted by the medical interview data transmission means 52 of terminal 10b, when executed. Server 10a executes a computer control program 50 to make a judgment such as a beauty assessment of the customer based on the medical interview data received by the medical interview data receiving means 54 and the analyzed parameters stored in its memory means. Server 10a's memory means stores a judgment result transmission means (subprogram) 55 for transmitting the judgment result, such as a beauty assessment, to terminal 10b when executed. The storage means of terminal 10b stores a judgment result receiving means (subprogram) 53 for receiving judgment results such as customer beauty judgments transmitted by the judgment result transmission means 55 of server 10a, and a judgment result output means (subprogram) 91b for outputting the received judgment results such as beauty judgments to the output device 40b.

[0114] In this embodiment, the server 10a may be configured to store customer medical interview data received by the medical interview data receiving means 54 in the storage means 30 of the server 10a by executing the computer control program 50, and when the number of stored medical interview data items etc. meets a predetermined standard, add all or part of those medical interview data to the basic data and store the updated basic data in the storage means 30, perform analysis such as factor analysis based on the basic data to update the analyzed parameters, and store the updated analyzed parameters 33 in the storage means 30.

[0115] <7-3. A system that performs beauty assessments and other evaluations on a terminal> Figure 14B is a diagram showing the configuration of a beauty assessment system in yet another embodiment of the present invention. Generally, factor analysis requires computationally intensive tasks, but if the analyzed parameters 33 of the model are known, the calculation of factor scores and the determination of skin type are generally computationally intensive and can be performed instantly on a mobile device or other terminal. Therefore, this embodiment is configured so that the calculation of factor scores and the determination of skin type are performed on terminal 10b.

[0116] In the beauty assessment system of this embodiment, in principle, in addition to inputting medical interview data and outputting assessment results such as beauty assessments, the terminal 10b performs calculations necessary for assessments such as beauty assessments, such as calculating factor scores and determining skin type. The terminal 10b in this embodiment comprises the basic configuration 11b (basic configuration A), an input device 20b, and an output device 40b. The terminal 10b further stores in the storage means 30 a medical interview data input means (subprogram) 51 for inputting customer medical interview data using the input device 20b and storing it in the storage means 30, and a medical interview data transmission means (subprogram) 52 for transmitting the medical interview data input by the medical interview data input means 51 and stored in the storage means 30 to the server 10a via the network 12. The server 10a comprises the basic configuration 11a. Upon execution, a stores in the storage means of server 10a a medical interview data receiving means (subprogram) 54 for receiving customer medical interview data transmitted by medical interview data transmission means 52. Server 10a stores in its storage means an analyzed parameter transmission means (subprogram) 57 for transmitting analyzed parameters 33, obtained by analyzing basic data using methods such as factor analysis and stored in its storage means, to terminal 10b via the network 12 upon execution. Terminal 10b stores in its storage means an analyzed parameter receiving means (subprogram) 56 for receiving the analyzed parameters 33 transmitted by the analyzed parameter transmission means 57 and storing them in the storage means 30. Terminal 10b executes the computer control program 50 to make a judgment, such as a beauty assessment, of the customer based on the input customer medical interview data and the analyzed parameters 33 received by the analyzed parameter receiving means 56. The storage means of terminal 10b stores a judgment result output means (subprogram) 91b for outputting the judgment results, such as the beauty judgment of the customer, to the output device 40b.

[0117] In this embodiment as well, the server 10a may be configured to store customer medical interview data received by the medical interview data receiving means 54 in the storage means of the server 10a by executing the computer control program 50, and when the number of stored medical interview data items etc. meets a predetermined standard, add all or part of those medical interview data to the basic data, store the updated basic data in the storage means, perform analysis such as factor analysis based on the basic data to update the analyzed parameters, and store the updated analyzed parameters 33 in the storage means.

[0118] The present invention is not limited to the embodiments and examples described above, and it goes without saying that various modifications and design changes within the scope of the present invention are included without departing from the technical spirit of the invention. [Industrial applicability]

[0119] According to the present invention, by using a factor analysis method based on big data to clarify the relationship between oily or dry skin types and self-perceived sensitive skin, it is possible to provide a beauty assessment system and a sensory evaluation system that can objectively and accurately determine a customer's skin type based on questionnaire data. According to the present invention, recommendations for cosmetics and beauty services, and sensory evaluations of beauty products can be performed more accurately than before in online or physical stores such as beauty salons, cosmetics stores, clinics, drugstores, and aesthetic salons. The present invention has broad industrial applicability. [Explanation of symbols]

[0120] 10 Computer 10a Server 10b Terminal 11 Basic Configuration A 12 Network 20 Input Devices 30 Storage means 31 Internal storage means 32 External storage means (storage medium) 33 Analyzed parameters 40 Output device 50 Computer control program 51 Medical interview data input means 52 Medical interview data transmission means 53. Means for receiving judgment results 54. Means for receiving medical interview data 55 Means for transmitting judgment results 56 Means for receiving analyzed parameters 57. Means for transmitting analyzed parameters 60. Beauty evaluation program 61 Skin type determination method 70 Dietary lifestyle analysis program 71 Dietary Habits Assessment Method 80 Health Condition Analysis Program 81. Means for determining physical condition 90. Result output program 91 Judgment result output means

Claims

1. It consists of independent computers, or multiple computers capable of bidirectional communication with each other. This beauty assessment system processes questionnaire data, which consists of customer responses to three or more questions about their skin, by executing a computer-controlled program, and performs a beauty assessment of the customer. The system has storage means for storing the customer's medical interview data, a computer control program, and factor score coefficients. The computer control program uses the customer's medical interview data and the factor score coefficients to calculate the customer's score for factor O related to excess sebum (SO), the score for dryness (SD) of factor D related to skin dryness, and the score for factor H related to pore dirt sensation (SH), and has a skin type determination means for determining the customer's skin type from the position of points (SO, SD, SH) in a skin type identification space equipped with XYZ coordinate axes. The factor score coefficients are pre-calculated by extracting at least the linearly independent O factor, D factor, and H factor from basic data consisting of responses from numerous subjects to the aforementioned questions regarding skin, using a factor analysis method. Beauty assessment system.

2. The beauty judgment system according to claim 1, wherein the aforementioned question includes at least one of the items related to pores, pore dirt, enlarged pores, and blackheads, which are closely related to the sensation of clogged pores.

3. Using the score SO of factor O, the score SH of factor H, and the angle t such that 0° < t < 90°, the combined oiliness SC is determined by the following formula 1: (Equation 1) SC = SO × cos(t) + SH × sin(t) The beauty determination system according to claim 2, wherein the skin type determination means calculates the customer's combined oiliness SC based on the medical interview data and determines the customer's skin type from the position of points (SC, SD) on a skin type identification plane having XY coordinate axes.

4. The computer control program has means for calculating the customer's or subject's skin quality score SQ using the customer's or subject's combined oiliness SC, dryness SD, and an angle s such that 0° < s < 90°, according to the following formula 2: (Equation 2) SQ = SC × cos(s) + SD × sin(s) The beauty judgment system according to claim 3, wherein the questions relating to the basic data include, separately from the questions relating to the answers used in the factor analysis, a question asking whether the subject is aware of their skin type as sensitive skin, and the angles t and s are determined such that the correlation coefficient R calculated from the distribution of the subject's skin type score SQ and the answers to the question on whether they are aware of their skin type as sensitive skin in the basic data is 0.90 or higher.

5. The aforementioned medical questionnaire data further includes the customer's answers to several questions regarding their eating habits, The aforementioned storage means further stores factor score coefficients related to dietary habits, The computer control program has means for calculating the score SF1 of the F1 factor related to irregular eating and the score SF2 of the F2 factor related to unhealthy eating for the customer, using the factor score coefficient related to dietary habits and the customer's medical interview data, and further calculating the score SF of the customer's combined dietary factors using the following formula 3, with an angle u such that 0° < u < 90°. (Formula 3) SF = SF1×cos(u)+SF2×sin(u) The aforementioned basic data further includes dietary basic data, which consists of the responses of numerous subjects to the aforementioned multiple questions regarding their eating habits. The factor score coefficients related to dietary habits are calculated in advance by extracting at least two linearly independent factors, F1 related to irregular eating and F2 related to unhealthy eating, from the basic dietary data using factor analysis. Using the factor score coefficients related to dietary habits, the score SF1 for factor F1 and the score SF2 for factor F2 for each subject in the basic dietary data are calculated in advance, and furthermore, the score SF for the combined dietary factors is calculated in advance using Equation 3. The beauty assessment system according to claim 4, wherein the angle u is determined such that the correlation coefficient R calculated from the distribution of the subject's score SF for the composite diet factor and the responses to the question regarding whether or not they perceive their skin as sensitive is 0.90 or greater in the basic data.

6. The aforementioned medical questionnaire data further includes the customer's responses to several questions regarding their perceived physical condition, The aforementioned memory means further stores factor score coefficients related to subjective perception of physical condition, The computer control program has means for calculating the T factor score ST related to the customer's physical condition specific to menopause, using the factor score coefficient related to the customer's subjective perception of physical condition and the customer's medical interview data. The aforementioned basic data further includes basic health data, which consists of responses from numerous subjects to the aforementioned multiple questions regarding their perceived physical condition. The factor score coefficients related to subjective perception of physical condition are calculated in advance by extracting at least two linearly independent factors, Factor A related to menopausal age and Factor T related to physical condition specific to menopause, from the aforementioned basic health data using factor analysis. Using the factor score coefficients related to subjective perception of physical condition, the T factor score ST for each subject in the basic health data is calculated in advance. The beauty assessment system according to claim 4, wherein the correlation coefficient R calculated from the distribution of the subject's T factor score ST and the responses to the question regarding whether or not they perceive themselves as having sensitive skin in the basic data is 0.85 or greater.

7. The computer control program has means for calculating the score SK of the circulatory factors related to the customer's circulatory problems using the factor score coefficient related to the customer's perception of their physical condition and the customer's medical interview data. The factor score coefficients related to subjective perception of physical condition are calculated by extracting at least circulatory factors related to circulatory dysfunction from the aforementioned basic health data using factor analysis. Using the factor score coefficients related to subjective perception of physical condition, the score SK for the circulatory factor for each subject in the basic health data is calculated in advance. The beauty assessment system according to claim 6, wherein the correlation coefficient R calculated from the distribution of the subject's score SK for the circulatory factor and the responses to the question regarding whether or not they perceive themselves as having sensitive skin in the basic data is 0.90 or higher.

8. The computer control program has means for calculating the customer's AL factor score (SAL) related to allergies, using the factor score coefficient related to the customer's perceived physical condition and the customer's medical interview data. The factor score coefficient related to subjective health perception is calculated by extracting at least the AL factor related to allergies from the basic health data using factor analysis. Using the factor score coefficients related to subjective perception of physical condition, the AL factor score SAL for each subject in the basic health data is calculated in advance. The beauty assessment system according to claim 6, wherein the correlation coefficient R calculated from the distribution of the subject's score SAL for the AL factor and the responses to the question regarding whether or not they perceive themselves as having sensitive skin in the basic data is 0.85 or higher.

9. The beauty determination system according to any one of claims 4 to 8, wherein the angle t is 25° or more and 50° or less, and the angle s is 25° or more and 75° or less.

10. The computer control program according to claim 9.

11. A computer-readable storage medium that records parameters relating to the computer control program described in claim 9.

12. A sensory evaluation system for a beauty product comprising the beauty evaluation system described in claim 9, wherein a collaborator who cooperates in the evaluation uses the beauty product for a certain period of time, and the collaborator is interviewed both before and after the collaborator uses the beauty product for that period of time, and based on the collaborator's answers to the questions, one or more numerical values ​​from among the O factor score, H factor score, combined oiliness, dryness, and skin quality score are calculated both before and after the period, and the improvement effect is evaluated from the amount of change in the numerical values ​​during the period.