Food sensory evaluation method and system based on eye movement tracking

Abstract

The invention discloses a food sensory evaluation method and system based on eye movement tracking, and belongs to the field of sensory digitization. The method comprises the steps that a user performs sensory evaluation scoring on food to be detected, and eye movement data and a corresponding scoring result are obtained in the evaluation scoring process; processing and calculating the eye movement data and the corresponding scoring result, and screening the score with the longest watching time and the score finally selected by clicking in the process of evaluating and scoring each characteristic attribute of each sample by the user; respectively carrying out mean value calculation and principal component analysis on the score with the longest watching time for evaluating each characteristic attribute of each sample by each user and the corresponding scoring result; and visually outputting an eye dynamic evaluation result. According to the invention, through collecting, sorting and analyzing the watching observation condition of the evaluator on the marking table during sensory evaluation, a solution is provided for solving the technical problems of missing data, single evaluation result and complicated and difficult analysis of eye movement data in the process of a food sensory evaluation method.

Description

Technical Field

[0001] The present invention relates to the field of sensory digitization technology, and more particularly to a food sensory evaluation method and system based on eye tracking. Background Art

[0002] Food sensory evaluation mainly relies on self-report questionnaires based on the subjective judgment of tasters, and the evaluation criteria have large individual differences. This method is difficult to accurately capture the taster's real experience of food and is easily affected by emotions, environment and tasting experience, resulting in inconsistent evaluation results and difficulty in standardization.

[0003] As an objective biometric technology, eye tracking can record the visual focus and gaze time of tasters during the sensory experience, providing direct response data to sensory stimulation. However, there is currently no method to systematically apply eye tracking technology to food sensory evaluation, which results in technical problems such as missing data, single evaluation results, and complex eye movement data that is difficult to analyze.

[0004] Therefore, how to provide a food sensory evaluation method and system based on eye tracking that can solve the above problems is an issue that technical personnel in this field urgently need to solve. Summary of the invention

[0005] In view of this, the present invention provides a food sensory evaluation method and system based on eye tracking, which collects, organizes and analyzes the gaze observation of the evaluator on the scoring scale during the sensory evaluation, so as to solve the technical problems of missing process data, single evaluation results and complex eye movement data that are difficult to analyze in the food sensory evaluation method.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A food sensory evaluation method based on eye tracking, comprising:

[0008] S100: The user performs sensory evaluation and scoring on the food to be tested, and obtains eye movement data and corresponding scoring results during the evaluation and scoring process;

[0009] S200: Processing and calculating eye movement data and corresponding scoring results, screening the score of the longest gaze time and the final click selection score in the process of evaluating and scoring each characteristic attribute of each sample by the user;

[0010] S300: performing mean calculation and principal component analysis on the longest gaze time of each user for each characteristic attribute of each sample and the corresponding scoring result;

[0011] S400: Visually output the eye movement sensory evaluation results.

[0012] Optionally, the user performs sensory evaluation and scoring on the food to be tested based on the constructed sensory evaluation and scoring system, and the sensory evaluation and scoring system construction process is:

[0013] Obtaining sensory characteristic attribute words of the food to be tested, and constructing a sensory characteristic attribute word system of the food to be tested;

[0014] The sensory characteristic attribute word system is used to construct a sensory evaluation scoring system.

[0015] Optionally, obtaining sensory characteristic attribute words of the food to be tested and constructing a sensory characteristic attribute word system of the food to be tested, including:

[0016] The review text of the food to be tested is used as the source text data, and the source text data is tokenized;

[0017] Perform stemming and lemmatizing.

[0018] The proper nouns are removed through Named Entity Removal;

[0019] Confirm the part of speech of each word and filter out nouns, adjectives, and verbs;

[0020] Perform word sense disambiguation in NLP and unify synonyms;

[0021] Perform word frequency screening to form the final sensory characteristic attribute word system of the food to be tested.

[0022] Optionally, a sensory evaluation scoring system is constructed using the sensory characteristic attribute word system, including:

[0023] Based on the sensory attribute word system of the food to be tested, the guiding words and the corresponding scoring scale are obtained;

[0024] Compile the guide words and scoring scale into a sensory evaluation scoring system;

[0025] The page of the sensory evaluation scoring system is adjusted to obtain a constructed sensory evaluation scoring system.

[0026] Optionally, adjust the sensory evaluation scoring system page, including:

[0027] Set the background size to 1920×1080 pixels;

[0028] Set the background color to light gray RGB(204, 204, 204);

[0029] Adjust the distribution of each page to keep it uniform and easy to read.

[0030] Optionally, the eye movement data and the corresponding scoring results are processed and calculated, and the score with the longest gaze time and the score of the final click selection are screened during the user's evaluation and scoring of each characteristic attribute of each sample, including:

[0031] After all users have completed the evaluation and scoring, the interest areas are drawn and named on the characteristic attribute words and corresponding score positions in the evaluation and scoring results of different users to distinguish them, and a database file is generated;

[0032] Integrate and calculate the generated database files to obtain the behavior indicator data of each user in each score interest area;

[0033] Filter the data corresponding to the score interest area whose total fixation duration is greater than 0 or whose click time is not empty in the behavior indicator data to obtain the filtering result;

[0034] According to the above screening results, the score with the longest total gaze time and the largest click time in the process of each user's evaluation and scoring of each characteristic attribute of each food to be tested are screened.

[0035] Optionally, the behavior indicators include: total fixation time TFL and last click time LCT.

[0036] A food sensory evaluation system based on eye tracking, comprising:

[0037] Data acquisition module: Users conduct sensory evaluation and scoring of the food to be tested, and obtain eye movement data and corresponding scoring results during the evaluation and scoring process;

[0038] Data screening module: processes and calculates eye movement data and corresponding scoring results, and screens the longest gaze time and the final click score of the user in the process of evaluating and scoring each characteristic attribute of each sample;

[0039] Data processing module: perform mean calculation and principal component analysis on the longest gaze time of each user for each characteristic attribute of each sample and the corresponding scoring results;

[0040] Visualization output module: Visualize the output of eye movement sensory evaluation results.

[0041] Optionally, it also includes a construction module for constructing a sensory characteristic attribute word system for the food to be tested, and on this basis constructing a sensory evaluation scoring system.

[0042] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a food sensory evaluation method and system based on eye movement tracking, in which users perform sensory evaluation and scoring on the food to be tested based on the constructed sensory evaluation scoring system, and combine the eye movement data in the scoring process to solve the problems of strong subjectivity and poor data consistency in the traditional questionnaire method, improve the objectivity of food sensory evaluation, and reduce the influence of human factors; obtain the user's (consumer's) real sensory experience of food samples through quantified eye movement data; provide a systematic processing solution for the big data of food sensory evaluation results; combine eye movement data and the user's constructed sensory evaluation scoring system to perform sensory evaluation and scoring feedback on the food to be tested, establish a correlation model between consumer preferences and food sensory characteristics, and provide a scientific basis for food product development. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0045] Figure 2 It is a schematic diagram of the system structure of the present invention;

[0046] Figure 3 A schematic diagram of the guide words and scoring scale page for building a sensory evaluation scoring system for coffee drinks;

[0047] Figure 4 Schematic diagram of the initial sight line page for calibration of the evaluator before entering the evaluation experiment;

[0048] Figure 5 Schematic diagram of the first page for building a sensory evaluation scoring system for coffee drinks;

[0049] Figure 6 Schematic diagram for initial parameter setting of eye tracking system;

[0050] Figure 7 This is a radar chart of the longest mean score of the eye movement-questionnaire sensory evaluation of coffee drinks;

[0051] Figure 8 For the eye movement-questionnaire sensory evaluation of coffee drinks, finally click to select the radar chart of the mean score;

[0052] Fig. 9 This is the principal component analysis diagram of the longest gaze time mean score of the eye movement-questionnaire sensory evaluation of coffee drinks;

[0053] Fig.10 Click to select the principal component analysis chart of the mean score for the eye movement-questionnaire sensory evaluation of coffee drinks. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] The purpose of the present invention is to provide a food sensory evaluation method and system based on eye movement tracking, the method comprising: S100: the user performs sensory evaluation and scoring on the food to be tested, and obtains eye movement data and corresponding scoring results during the evaluation and scoring process; S200: processing and calculating the eye movement data and the corresponding scoring results, screening the score with the longest gaze time and the score finally clicked and selected during the evaluation and scoring process of each characteristic attribute of each sample by the user; S300: performing mean calculation and principal component analysis on the score with the longest gaze time for each characteristic attribute of each sample evaluated by each user and the corresponding scoring result; S400: visually outputting the eye movement sensory evaluation results. The present invention collects, organizes and analyzes the gaze observation of the evaluator on the scoring scale during the sensory evaluation, and provides a solution to the technical problems of missing process data, single evaluation results, and complicated and difficult to analyze eye movement data in the food sensory evaluation method.

[0056] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] Embodiment 1:

[0058] See also Figure 1 The embodiment of the present invention discloses a food sensory evaluation method based on eye tracking, comprising:

[0059] S100: The user performs sensory evaluation and scoring on the food to be tested, and obtains eye movement data and corresponding scoring results during the evaluation and scoring process;

[0060] S200: Processing and calculating eye movement data and corresponding scoring results, screening the score of the longest gaze time and the final click selection score in the process of evaluating and scoring each characteristic attribute of each sample by the user;

[0061] S300: performing mean calculation and principal component analysis on the longest gaze time of each user for each characteristic attribute of each sample and the corresponding scoring result;

[0062] S400: Visually output the eye movement sensory evaluation results.

[0063] In a specific embodiment, the user performs sensory evaluation and scoring on the food to be tested based on the constructed sensory evaluation and scoring system, and the sensory evaluation and scoring system construction process is:

[0064] Obtaining sensory characteristic attribute words of the food to be tested, and constructing a sensory characteristic attribute word system of the food to be tested;

[0065] A sensory evaluation scoring system is constructed using the sensory characteristic attribute word system of the food to be tested.

[0066] In a specific embodiment, obtaining sensory characteristic attribute words of the food to be tested and constructing a sensory characteristic attribute word system of the food to be tested includes:

[0067] The review text of the food to be tested is used as the source text data, and the source text data is tokenized;

[0068] Perform stemming and lemmatizing.

[0069] The proper nouns are removed through Named Entity Removal;

[0070] Confirm the part of speech of each word and filter out nouns, adjectives, and verbs;

[0071] Perform word sense disambiguation in NLP and unify synonyms;

[0072] Perform word frequency screening to form the final sensory characteristic attribute word system of the food to be tested.

[0073] In a specific embodiment, the sensory characteristic attribute word system is used to construct a sensory evaluation scoring system, including:

[0074] Based on the sensory attribute word system of the food to be tested, the guiding words and the corresponding scoring scale are obtained;

[0075] Compile the guide words and scoring scale into a sensory evaluation scoring system;

[0076] The page of the sensory evaluation scoring system is adjusted to obtain a constructed sensory evaluation scoring system.

[0077] In a specific embodiment, adjusting the page of the sensory evaluation scoring system includes:

[0078] Set the background size to 1920×1080 pixels;

[0079] Set the background color to light gray RGB(204, 204, 204);

[0080] Keep each page evenly distributed and easy to read.

[0081] In a specific embodiment, the eye movement data and the corresponding scoring results are processed and calculated, and the score of the longest gaze time and the score of the final click selection by the user in the process of evaluating and scoring each characteristic attribute of each sample are screened, including:

[0082] After all users have completed the evaluation and scoring, the interest areas are drawn and named on the characteristic attribute words and corresponding score positions in the evaluation and scoring results of different users to distinguish them, and a database file is generated;

[0083] Integrate and calculate the generated database files to obtain the behavior indicator data of each user in each score interest area;

[0084] Filter the data corresponding to the score interest area whose total fixation duration is greater than 0 or whose click time is not empty in the behavior indicator data to obtain the filtering result;

[0085] According to the above screening results, the score with the longest total gaze time and the largest click time in the process of each user's evaluation and scoring of each characteristic attribute of each food to be tested are screened.

[0086] In a specific embodiment, the behavior indicators include: total fixation time (TFL) and last click time (LCT).

[0087] Specifically, in this embodiment, when processing and calculating eye movement data and corresponding scoring results, the processing and calculation are to calculate the behavior indicators within the scoring AOI, including the scoring AOI gaze duration and click time. The eye movement data records the user's observation of the scoring scale during the scoring process; the scoring result refers to the final click selection result.

[0088] Specifically, each characteristic attribute refers to a characteristic attribute used to describe the food sample to be evaluated, which comes from the sensory dictionary of the food to be evaluated, or an evaluation panel, or similar research articles.

[0089] In a specific embodiment, the longest gaze time score of each user for each characteristic attribute of each sample and the corresponding scoring result are respectively averaged and subjected to principal component analysis, including:

[0090] The average of the longest gaze time score and the final click score of each characteristic attribute of each sample evaluated by the evaluator is calculated, and the result is visualized as a radar chart.

[0091] In a specific embodiment, the average of the scores of the longest gaze time of the evaluators on each characteristic attribute of each sample is calculated, and the average of the scores of the final click selections of the evaluators on each characteristic attribute of each coffee beverage sample is calculated, and the results are visualized as radar charts respectively.

[0092] Furthermore, principal component analysis was performed on the sample-feature attribute-longest gaze score average matrix and the sample-feature attribute-final click selection score average matrix. The principal component analysis was performed with the sample as the observation value and the feature attribute as the loading to obtain the principal component analysis results.

[0093] Specifically, the sample-feature attribute-longest fixation score average matrix and the sample-feature attribute-final click selection score average matrix are both sample×feature attribute matrices, where each row represents a sample and each column represents a feature attribute. Each element of the matrix is ​​the average score of the corresponding sample and feature attribute.

[0094] Taking the longest gaze score average matrix as an example, for coffee sample A, its characteristic attributes include sour, sweet, salty, and bitter. The sample A row will contain 4 elements, which represent the average values ​​of the longest gaze time of the evaluators when evaluating the sour, sweet, salty, and bitter tastes of sample A.

[0095] Specifically, mean calculation and principal component analysis are basic statistical methods. The specific algorithm content refers to the implementation by code, and the calculation logic is:

[0096] Mean calculation is to calculate the average of the scores obtained by the two methods.

[0097] The specific steps of principal component analysis are:

[0098] Data standardization: Eliminate dimensional differences.

[0099] Calculate the covariance matrix: measure the correlation between feature attributes.

[0100] Solving for eigenvalues ​​and eigenvectors: Eigenvectors represent the principal directions of the data, while eigenvalues ​​represent the importance of these directions.

[0101] Select principal components: Select the two eigenvectors with the largest contribution as principal components based on the size of the eigenvalues.

[0102] Construct principal component scores: Project the original data onto the selected principal components to obtain principal component scores, which can be used for subsequent analysis or modeling.

[0103] In a specific embodiment, the embodiment of the present invention provides the steps of principal component analysis in statistics. In actual operation, the PCA function in the sklearn library in Python is used for analysis. The code is as follows:

[0104] pca = PCA(n_components = 2)

[0105] principal_components=pca.fit_transform(pivot_table_standardized.fillna(0)).

[0106] See also Figure 2 The embodiment of the present invention further discloses a food sensory evaluation system based on eye tracking, comprising:

[0107] Data acquisition module: Users conduct sensory evaluation and scoring of the food to be tested, and obtain eye movement data and corresponding scoring results during the evaluation and scoring process;

[0108] Data screening module: processes and calculates eye movement data and corresponding scoring results, and screens the longest gaze time and the final click score of the user in the process of evaluating and scoring each characteristic attribute of each sample;

[0109] Data processing module: perform mean calculation and principal component analysis on the longest gaze time of each user for each characteristic attribute of each sample and the corresponding scoring results;

[0110] Visualization output module: Visualize the output of eye movement sensory evaluation results.

[0111] Specifically, it also includes a construction module for constructing a sensory characteristic attribute word system for the food to be tested, and on this basis, constructing a sensory evaluation scoring system.

[0112] Embodiment 2:

[0113] This embodiment further illustrates a food sensory evaluation method and system based on eye tracking provided by the present invention in conjunction with a sensory evaluation of coffee beverages.

[0114] On the one hand, an embodiment of the present invention provides a food sensory evaluation method based on eye tracking, which performs sensory evaluation on coffee beverages, comprising the following steps:

[0115] First, we build a coffee sensory attribute system. We use website reviews of coffee drinks as source text data, tokenize the review text, and then perform stemming and lemmatizing. We then remove proper nouns through named entity removal, confirm the part of speech of each word, and filter out nouns, adjectives, and verbs (POS (part of speech) filtering). Then, we perform word sense disambiguation in NLP, that is, synonym unification, and finally perform frequency filtering to form the final coffee drink attribute word system.

[0116] In a specific embodiment, the characteristic attribute words can also be selected from the four aspects of vision, smell, taste, and touch with reference to the "Coffee Sensory Dictionary" to form a characteristic attribute word list and prepare a CATA questionnaire. The evaluator makes multiple selections on the characteristic attribute words suitable for describing the target food in the CATA questionnaire based on the characteristic attributes that coffee drinks should have. The characteristic attribute words with a selection frequency greater than 80% are selected as the characteristic attribute word system for coffee drinks.

[0117] The evaluator can also freely associate the characteristic attribute words of coffee drinks from the four aspects of vision, smell, taste and touch, collect them using questionnaires, visualize the results with the help of word clouds, and then select characteristic attribute words to form a list of characteristic attribute words to make a CATA questionnaire. The evaluator makes multiple choices on the characteristic attribute words suitable for describing the target food in the CATA questionnaire based on the characteristic attributes that coffee drinks should have. The characteristic attribute words with a selection frequency greater than 80% are selected as the characteristic attribute word system of coffee drinks.

[0118] In a specific embodiment, constructing a sensory evaluation scoring system for coffee beverages specifically includes:

[0119] The Eyeso studio + Eyeso Ec80 eye tracking system from Xintuo Yingqi was used, and its initial parameter settings were as follows Figure 6 The guide words and the rating scale questionnaire are edited into the eye tracking system as stimulus materials. The guide words and the rating scale are as shown in FIG. Figure 3 As shown in Figure 2, the central cross image is used to calibrate the initial sight line of the evaluator before entering the evaluation experiment. Figure 4 As shown; Figure 5The first page of the sensory evaluation scoring system for entering coffee drinks is constructed. In particular, in this embodiment, in order to adapt to the display resolution of the eye tracker, the background size is set to 1920×1080 pixels; at the same time, in order to avoid the background color and brightness from interfering with the eye movement behavior of the participants, the background color is set to light gray RGB (204, 204, 204). In addition, in this embodiment, in order to ensure the uniformity and readability of the distribution of the scoring scale, its overall layout is set as follows: each slide contains a maximum of 4 sensory characteristic attributes, each standard is arranged vertically, with a spacing of 192 pixels (2 inches) between each other; there are 9 scoring options (1 to 9) next to each standard, arranged horizontally, with a spacing of 144 pixels (1.5 inches) between each option.

[0120] In a specific embodiment, 33 consumer evaluators are organized, configured and presented with coffee beverage samples. The evaluators sit in front of the eye tracking device, place their heads on the headrest, and ensure that the distance between their eyes and the screen is about 60 cm. The evaluators first calibrate the eye tracking device, and then start the sensory evaluation eye movement-questionnaire experiment record. The evaluators need to observe, smell or taste the coffee samples, and complete the click scoring on the scoring scale. During the evaluation and scoring process, the gaze observation data is obtained at the same time to obtain the eye movement-questionnaire sensory evaluation scoring results.

[0121] In a specific embodiment, the eye movement-questionnaire sensory evaluation scoring results of coffee beverages are processed and calculated: after all evaluators complete the evaluation experiment, interest areas are drawn at the characteristic attribute words and their score positions and named to distinguish them. The AOI name is named in the format of "(characteristic attribute word)" or "(characteristic attribute word) (score)", such as "acid" or "acid 1"; the result database files of the evaluation experiments on different computers are integrated and calculated to obtain various behavioral indicators of each evaluator in each score interest area: total fixation time (TFL), last click time (LCT). In the process of evaluating and scoring each characteristic attribute of each coffee beverage sample, the evaluator is screened, and the score with the longest total fixation time and the score of the final click selection are selected. Specifically, the prerequisite for the selection of the two scores is that the total fixation time is greater than 300ms.

[0122] In a specific embodiment, the longest gaze time score and the final click score of each characteristic attribute of each coffee beverage sample evaluated by the evaluator are averaged, and the result is visualized as a radar chart, such as Figure 7-Figure 8 , radar chart of the mean values ​​of eye movement-questionnaire sensory evaluation of coffee drinks.

[0123] In a specific embodiment, the average value of the longest gaze time of the evaluators for each characteristic attribute of each coffee beverage sample is calculated, and the average value of the final click selection score of the evaluators for each characteristic attribute of each coffee beverage sample is calculated, and the results are visualized as radar charts, such as Figure 7-Figure 8 .

[0124] Furthermore, principal component analysis was performed on the coffee beverage sample-feature attribute longest gaze score average matrix and the coffee beverage sample-feature attribute final click selection score average matrix. The principal component analysis results are shown in the figure. Figure 9-10 shown.

[0125] In a specific embodiment, the eye movement-questionnaire sensory evaluation analysis results of the coffee beverage are visualized and output.

[0126] On the other hand, the present invention provides a food sensory evaluation system based on eye tracking, which performs sensory evaluation on coffee beverages, including:

[0127] The construction module is used to construct a sensory characteristic attribute word system for the food to be tested, and on this basis, a sensory evaluation scoring system is constructed.

[0128] A data acquisition module, connected to the construction module, is used to obtain the sensory evaluation and scoring data of the food to be tested based on eye tracking by the evaluator, and simultaneously obtain the gaze observation data during the evaluation and scoring process to obtain the eye movement-questionnaire sensory evaluation and scoring results;

[0129] Data screening module: connected to the acquisition module, processes and calculates eye movement data and corresponding scoring results, and screens the longest gaze time and the final click score of the evaluator in the process of evaluating and scoring each characteristic attribute of each sample;

[0130] Specifically, the data screening module includes the following sub-modules: database integration module, AOI identification module, index calculation module, and score screening module;

[0131] A data processing module, connected to the data screening module, is used to calculate the mean of the longest gaze time score and the final click score of each evaluator for each characteristic attribute of each sample, and further perform principal component analysis;

[0132] Visualization output module: used to output and display the sensory evaluation results of the eye movement questionnaire.

[0133] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0134] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A food sensory evaluation method based on eye tracking, characterized in that: include: S100: The user performs sensory evaluation and scoring on the food to be tested, and obtains eye movement data and corresponding scoring results during the evaluation and scoring process; S200: Processing and calculating eye movement data and corresponding scoring results, screening the score of the longest gaze time and the final click selection score in the process of evaluating and scoring each characteristic attribute of each sample by the user; S300: performing mean calculation and principal component analysis on the longest gaze time of each user for each characteristic attribute of each sample and the corresponding scoring result; S400: Visually output the eye movement sensory evaluation results.

2. The food sensory evaluation method based on eye tracking according to claim 1, characterized in that: The user performs sensory evaluation and scoring on the food to be tested based on the constructed sensory evaluation and scoring system. The construction process of the sensory evaluation and scoring system is as follows: Obtaining sensory characteristic attribute words of the food to be tested, and constructing a sensory characteristic attribute word system of the food to be tested; A sensory evaluation scoring system is constructed using the sensory characteristic attribute word system of the food to be tested.

3. The food sensory evaluation method based on eye tracking according to claim 2, characterized in that: Obtain the sensory characteristic attribute words of the food to be tested, and construct a sensory characteristic attribute word system of the food to be tested, including: The review text of the food to be tested is used as the source text data, and the source text data is tokenized; Perform stemming and lemmatizing. The proper nouns are removed through Named Entity Removal; Confirm the part of speech of each word and filter out nouns, adjectives, and verbs; Perform word sense disambiguation in NLP and unify synonyms; Conduct word frequency screening to form the final sensory characteristic attribute word system of the food to be tested.

4. The food sensory evaluation method based on eye tracking according to claim 2, characterized in that: The sensory characteristic attribute word system is used to construct a sensory evaluation scoring system, including: Based on the sensory attribute word system of the food to be tested, the guiding words and the corresponding scoring scale are obtained; Compile the guide words and scoring scale into a sensory evaluation scoring system; The page of the sensory evaluation scoring system is adjusted to obtain a constructed sensory evaluation scoring system.

5. The food sensory evaluation method based on eye tracking according to claim 4, characterized in that: Adjust the sensory evaluation scoring system page, including: Set the background size to 1920×1080 pixels; Set the background color to light gray RGB(204, 204, 204); Adjust each page distribution.

6. The food sensory evaluation method based on eye tracking according to claim 1, characterized in that: Process and calculate the eye movement data and the corresponding scoring results, and filter the longest gaze time and the final click score of the user in the process of evaluating and scoring each characteristic attribute of each sample, including: After all users have completed the evaluation and scoring, the interest areas are drawn and named on the characteristic attribute words and corresponding score positions in the evaluation and scoring results of different users to distinguish them, and a database file is generated; Integrate and calculate the generated database files to obtain the behavior indicator data of each user in each score interest area; Filter the data corresponding to the score interest area whose total fixation duration is greater than 0 or whose click time is not empty in the behavior indicator data to obtain the filtering result; According to the above screening results, the score with the longest total gaze time and the largest click time in the process of each user's evaluation and scoring of each characteristic attribute of each food to be tested are screened.

7. The food sensory evaluation method based on eye tracking according to claim 6, characterized in that: The behavior indicators include: total fixation time TFL and click time CT.

8. A food sensory evaluation system based on eye tracking using the food sensory evaluation method based on eye tracking according to any one of claims 1 to 7, characterized in that: include: Data acquisition module: Users conduct sensory evaluation and scoring of the food to be tested, and obtain eye movement data and corresponding scoring results during the evaluation and scoring process; Data screening module: processes and calculates eye movement data and corresponding scoring results, and screens the longest gaze time and the final click score of the user in the process of evaluating and scoring each characteristic attribute of each sample; Data processing module: perform mean calculation and principal component analysis on the longest gaze time of each user for each characteristic attribute of each sample and the corresponding scoring results; Visualization output module: Visualize the output of eye movement sensory evaluation results.

9. The food sensory evaluation system based on eye tracking according to claim 8, characterized in that: It also includes a construction module for constructing a sensory characteristic attribute word system for the food to be tested, and on this basis, a sensory evaluation scoring system is constructed.

Images