Frontiers in Computing and Intelligent Systems ISSN: 2832-6024 | Vol. 13, No. 3, 2025 1 Study on Consumer Cognition Evaluation Model for Internet Celebrity Food Used by AHP Method Chenxuan Zhao 1, *, Qianmu Zhang 1, Lili Wang 1, Anjun Liu 2 1 Shandong Experimental High School, Jinan, Shandong province, 250001, China 2 Jinan Key Laboratory of High Performance Industrial Software, Jinan Institute of Supercomputing Technology, Jinan, Shandong province, 250100, China * Corresponding author: Chenxuan Zhao (Email: jnroni@126.com) Abstract: One mathematical model to quantitatively evaluate consumer recognition of internet celebrity food (ICF) was established, providing assistance in the development of such products. Firstly, through a questionnaire survey conducted among college students, combined with literature reports, an evaluation system was proposed comprising four topics of Product quality, Brand image, Marketing promotion, and Consumer experience, along with 14 specific indices. Secondly, the Analytic Hierarchy Process (AHP) method was employed to construct the evaluation model for consumer recognition of ICF, revealing that "Marketing promotion" was a crucial topic. Specifically, the comprehensive weight values of 14 indices were obtained. The indices of "Endorsement or Recommendation by Internet Influencers" and "Cultural Experience" with the high weight values reflected the internet popularity and cultural essence of ICF, underscoring their extreme importance in consumer recognition. Finally, the established model was tested by four kinds of ICF with high praise rates from the online shopping platform. And the results showed that the model had good stability to make quantitatively informed decisions about which food best fits the criteria of consumer recognition, which will provide a theoretical basis and practical reference for the development of ICF. Keywords: Internet Celebrity Food; Consumer Evaluation Model; Questionnaire Survey; AHP Method; Data Crawling. 1. Introduction Internet celebrity food (ICF) is a new type of product that emerged in the Chinese food industry during the e-commerce era, with examples including Luosifen (spiral rice noodle with chili sauce), Self-heating mini hotpots and Sauce-flavored lattes. With their unique concepts and novel appearances, these products could gain immense popularity and favor among numerous consumers [1-3], especially young people. Understanding the logic behind the immense success of ICF is very necessary and meaningful. Constructing a scientific and reliable consumer cognition evaluation model for ICF has important theoretical and practical significance, which can guide product development, marketing strategies, and quality supervision. Some studies have leveraged data mining and mathematical methods to decode consumer sentiments of ICF. Wang et al. [4] analyzed the visual communication design of interactive packaging of Internet-famous food using the weighted fusion rule of conjunction disjunction double operators. The innovative multi-group discrimination results of interactive packaging were obtained. Song et al. [5] analyzed 136835 online comments regarding Internet-famous food from Dianping platform between 2016 and 2019 using a web scraper. A sentiment lexicon for Internet-famous food was constructed, and sentiment analysis was further conducted to understand consumer attitudes. Huang et al. [6] used Structural Equation Modeling to establish a research model to investigate how interaction between online celebrity and consumers, which quantified how attractiveness, credibility, and expertise affect repurchase intentions. Fei et al. [7] established the model of ICF consumer trust formation by integrating the grounded theory and the decomposed theory of planned behavior. It was proposed that consumers followed the path of "pre-purchase factors-consumer perception-consumer trust formation-consumer behavior decision-making" in the process of online shopping for internet celebrity food. Regarding ICF in the context of the big data era, the previous researches have used mathematical model to preliminarily analyze consumers' perception of ICF. However, few studies have systematically analyzed consumer cognition and established a comprehensive evaluation model of ICF. This study will conduct a questionnaire survey among young college student consumers and analyze online comments from various platforms to explore the key factors that influenced consumers' attitudes towards ICF. One quantitative evaluation mathematical model is expected to be established to analyze the influencing mechanism of various factors, which affect consumer recognition. Furthermore, based on the analysis results, the improvement suggestions are provided to enhance consumer satisfaction, thereby promoting the development of ICF industry. 2. Materials and Methods 2.1. ICF Questionnaire Survey A questionnaire survey on internet celebrity foods was conducted among 98 college students from the School of Culture and Tourism, University of Jinan. The aim of the survey was to gain ICF insights into the key concerns and preferences of young consumers across five major dimensions: personal information and consumption habits, evaluations, consumption frequency and amount, and development perspectives. According to the survey results, specific indices related to consumer evaluations of ICF would be identified in order to establish an indices system. 2.2. AHP Method AHP (Analytic Hierarchy Process) method [8-10] was 2 utilized to construct a model for evaluating consumer recognition of ICF using the online version of SPSSPRO software. 5 professional teachers in the field of food and marketing were invited to judge the relative importance of the main criteria and sub-criteria. When comparing the criteria, the expert used a pairwise comparison scale containing the numerical values between 1 and 9 [11]. Maximum eigenvalue (λmax) was calculated as: 𝜆 ∑ (1) where matrix A represented the pairwise comparisons of the criteria, and the dimension n corresponded to the number of criteria being evaluated, while w represented the eigenvector. The consistency index (CI), which determined the acceptability of the consistency ratio for the comparison matrix A, was then calculated as: 𝐶𝐼 (2) To evaluate and reduce inconsistencies in the comparison matrix, the consistency ratio (CR) was calculated using the formula (3). If CR≤0.1, it indicated that the judgment matrix met the requirements. 𝐶𝑅 (3) where RI was the random index that varied according to the size of the matrix. 2.3. Test Sample Selection Utilizing the data acquisition method of focused web crawlers to obtain data on ICF provided a more objective approach compared to data collection through interviews and questionnaires. Based on the indices of the evaluation system for consumer recognition of ICF, the Octopus crawler software network version was employed to collect and organize the latest reviews on four best-selling ICF on the JD.COM platform [12]. All reviews used was up-to-date, spanning from April 1, 2024, to January 25, 2025, to ensure the timeliness of the research. Online reviews were obtained from the platform, and 4 kinds of ICF with over 2 million reviews and over 200000 positive reviews were selected. Total of 4000 review data entries were crawled and indices statistics were compiled for the testing of evaluation model. 3. Results 3.1. Evaluation Indices System Construction Drawing on questionnaire survey and the previous studies [13-15], 14 specific indices related to consumer evaluations of ICF were selected. These included Ingredient quality (C1), Taste and flavor (C2), Nutritional health (C3), Brand awareness (C4), Brand reputation (C5), Brand innovation (C6), Social media popularity (C7), Internet celebrity endorsement/recommendation (C8), Price (C9), Marketing campaign effectiveness (C10), Purchase convenience (C11), Packaging design (C12), Cultural experience (C13) and After- sales service (C14). These indices could be grouped into four topics for consideration: Product quality (B1), Brand image (B2), Marketing promotion (B3) and Consumption experience (B4). The above indices were organized to form a preliminary evaluation indices system for consumer recognition of ICF. This system was structured into three levels: the target level, the topic level and the indice level. The topic level was divided into four evaluation perspectives, and the indice level summarized 14 evaluation indices. Specifically, it could be summarized as shown in Figure 1. Figure 1. Evaluation indices system 3 3.2. Establishment of ICF Consumer Recognition Model 3.2.1. Judgment Matrix Construction and Weight Calculation of The Topic Layer Table 1. Topic level judgment matrix Topics Product Quality (B1) Brand Image (B2) Marketing and Promotion (B3) Consumption Experience (B4) Product Quality (B1) 1 3 0.2 0.25 Brand Image (B2) 0.333 1 0.143 0.2 Marketing and Promotion (B3) 5 7 1 3 Consumption Experience (B4) 4 5 0.333 1 The topic layer included Product Quality, Brand Image, Marketing and Promotion, and Consumption Experience. Under the framework of the AHP, a scoring matrix was constructed based on the judgment of 5 experts as shown in Table 1. Weight values of each topic presented clear differences in Table 2, reflecting the their degree of importance in the comprehensive evaluation. Marketing and Promotion had the highest corresponding weight value of 54.803%, which indicated that it was regarded as the most critical factors in the decision-making process, exerting a decisive influence on the overall evaluation. Closely following was Consumption Experience, with weight value of 28.079%, indicating that it also occupied an important position in the evaluation system and was a non-negligible aspect influencing decision-making. Table 2. Analysis results of AHP Topics Characteristic Vectors Weight Values (%) Largest Eigenvalue CI Product Quality (B1) 0.46 11.494 4.171 0.057 Brand Image (B2) 0.225 5.623 Marketing and Promotion (B3) 2.192 54.803 Consumption Experience (B4) 1.123 28.079 The consistency test result of the AHP was shown in Table 3. Judging from the provided data, the results of the consistency test with the maximum eigenvalue λmax being 4.171, the CI (Consistency Index) value being 0.057, the RI (Random Index) value being 0.88, and the CR (Consistency Ratio) value being 0.065 was "Pass". Table 3. Consistency test results λmax CI RI CR Results of Consistency Test 4.171 0.057 0.882 0.065 Pass 3.2.2. Judgment Matrix Construction and Weights Calculation of the Indice Level (1) Product Quality(B1) Based on scores provided by the experts, the judgment matrix for the B1 indices level was constructed as shown in Table 4. The weight values were calculated based on the eigenvectors and represented the proportion of each factor in the overall decision-making process. In Table 5, it could be seen that Taste and Flavor had the highest weight value (60.247%), occupying a dominant position. Nutrition and Health followed with a weight value of 31.512%, while Ingredients and Raw Materials had the lowest weight value (8.24%). This result reflected that in decision-making, Taste and Flavor were considered the most critical factors, followed by Nutrition and Health, while Ingredients and Raw Materials were relatively less important. The CR value was 0.002 in Table 6, which was far below the threshold of 0.1, indicating that the judgment matrix had a high degree of consistency and passed the consistency test. Table 4. Judgment matrix (B1) Indice Ingredients and Raw Materials (C1) Taste and Flavor (C2) Nutrition and Health (C3) Ingredients and Raw Materials (C1) 1 0.143 0.25 Taste and Flavor (C2) 7 1 2 Nutrition and Health (C3) 4 0.5 1 Table 5. Results of AHP hierarchical analysis (B1) Indice Characteristic Vectors Weight Values (%) Largest Eigenvalue CI Ingredients and Raw Materials (C1) 0.247 8.24 3.002 0.001 Taste and Flavor (C2) 1.807 60.247 Nutrition and health (C3) 0.945 31.512 Table 6. Consistency test results (B1) λmax CI RI CR Results of Consistency Test 3.002 0.001 0.525 0.002 Pass (2) Brand Image(B2) Table 7. Judgment matrix (B2) Indice Brand Awareness (C4) Brand Reputation (C5) Brand Innovation (C6) Brand Awareness (C4) 1 0.333 0.125 Brand Reputation (C5) 3 1 0.2 Brand Innovation (C6) 8 5 1 The judgment matrix for the B2 indices level was constructed as shown in Table 7. In Table 8, the data showed that the eigenvector for Brand Innovation was 2.211, with a corresponding weight value of 73.695%, indicating that Brand Innovation held the highest importance among all factors considered. Closely following was Brand Reputation, with an eigenvector of 0.559 and a weight value of 18.628%, suggesting that it was also a key factor. In contrast, Brand Awareness had a lower level of importance, with an 4 eigenvector of 0.23 and a weight value of 7.677%. The calculation results of AHP showed that the maximum eigenvalue was 3.044, and the Consistency Ratio (CR) was 0.042 in Table 9, which was less than 0.1, passing the one- time test. Table 8. Results of AHP hierarchical analysis (B2) Indice Characteristic Vectors Weight Values (%) Largest Eigenvalue CI Brand Awareness (C4) 0.23 7.677 3.044 0.022 Brand Reputation (C5) 0.559 18.628 Brand Innovation (C6) 2.211 73.695 Table 9. Consistency test results (B2) λmax CI RI CR Results of Consistency Test 3.044 0.022 0.525 0.042 Pass (3) Marketing and Promotion(B3) Table 10 showed the judgment matric for the B3. In table 11, the calculation results showed that the eigenvector for Endorsement or Recommendation by Internet Influencers was 2.231, with a weight value as high as 55.766%, which was the highest among all evaluated items, highlighting its central role in the evaluation system. This result reflected the high level of trust consumers placed in the opinions of internet influencers in the current market environment, and their recommendations were crucial for the acceptance of products or services. The eigenvector for Social Media Popularity was 1.000, with a corresponding weight value of 25.944%, indicating that it occupied a relatively important position in the system. The eigenvector for Effectiveness of Marketing Activities was 0.282, with a value of 7.053%, which was the lowest weight among the four evaluations. The calculation results of AHP showed that the maximum eigenvalue was 3.044, with a CR of 0.03 in Table 12, which was less than 0.1, passing the one-time test. Table 10. Judgment matrix (B3) Indice Social Media Popularity (C7) Endorsement or Recommendation by Internet Influencers (C8) Price (C9) Effectiveness of Marketing Activities (C10) Social Media Popularity (C7) 1 0.333 3 4 Endorsement or Recommendation by Internet Influencers (C8) 3 1 5 6 Price (C9) 0.333 0.2 1 2 Effectiveness of Marketing Activities (C10) 0.25 0.167 0.5 1 Table 11. Results of AHP hierarchical analysis (B3) Indice Characteristic Vectors Weight Values (%) Largest Eigenvalue CI Social Media Popularity (C7) 1.038 25.944 4.079 0.026 Endorsement or Recommendation by Internet Influencers (C8) 2.231 55.766 Price (C9) 0.449 11.237 Effectiveness of Marketing Activities (C10) 0.282 7.053 Table 12. Consistency test results (B3) λmax CI RI CR Results of Consistency Test 4.079 0.026 0.882 0.03 Pass (4) Consumption Experience(B4) Table 13-15 presented the judgment matrix, AHP analysis and consistency test results, respectively. The AHP results indicated that Cultural Experience significantly led with an eigenvector of 2.404 and a weight value of 60.108%, suggesting that Cultural Experience was the most central influencing factor in consumers' considerations and played a decisive role in the overall evaluation. Closely following was Purchase Convenience, with an eigenvector of 1.006 and a weight of 25.149%, indicating that purchase convenience was also an important factor that could not be ignored. In contrast, Packaging Design and After-sales Service had lower weight values of 5.594% and 9.148%, respectively, indicating that their influence was relatively small. The calculation results of AHP showed that the maximum eigenvalue was 3.044, with a CR of 0.052, which was less than 0.1, passing the one-time test. Table 13. Judgment matrix (B4) Indice Purchase Convenience (C11) Packaging Design (C12) Cultural Experience (C13) After-sales Service (C14) Purchase Convenience (C11) 1 5 0.25 4 Packaging Design (C12) 0.2 1 0.125 0.5 Cultural Experience (C13) 4 8 1 6 After-sales Service (C14) 0.25 2 0.167 1 Table 14. Results of AHP hierarchical analysis (B4) Indice Characteristic Vectors Weight Values (%) Largest Eigenvalue CI Purchase Convenience (C11) 1.006 25.149 4.139 0.046 Packaging Design (C12) 0.224 5.594 Cultural Experience (C13) 2.404 60.108 After-sales Service (C14) 0.366 9.148 5 Table 15. Consistency test results (B4) λmax CI RI CR Results of Consistency Test 4.139 0.046 0.882 0.052 Pass 3.2.3. Comprehensive Indices Weight Ranking and Analysis Based on the aforementioned algorithm, the weight values of each indices were calculated and comprehensively ranked for both the topic layer and the indice layer as shown in Table 16. Three indices of Endorsement or Recommendation by Internet Influencers, Cultural Experience and Social Media Popularity had comprehensive weight values above 14%, playing a crucial role in the development of ICF. Any weakness in any of these areas would inevitably have impacted the long-term development of such foods. Among them, Endorsement or Recommendation by Internet Influencers had a weight of 30.560, reflecting the internet popularity attribute and highlighting its extreme importance in consumer recognition of ICF. Cultural Experience ranked second and was the most intrinsic reason for the sustained popularity of ICF. Compared to the above strong influence indices, four weak influence indices of Brand Awareness, Ingredients and Raw Materials, Brand Reputation and Packaging Design had comprehensive weight values ranging from 0.430 to 1.570. Their role in the recognition of ICF were also relatively weak due to the relatively low values. And they should not have been the focus of future attention. Table 16. Weight ranking of indices in the consumer recognition evaluation system Target layer Topic layer Weight value(%) Sort Indice layer Weight value(%) Comprehensive weight(%) Sort Comprehensive Evaluation of Internet-famous Food (A) Product Quality (B1) 11.49 3 Ingredients and Raw Materials (C1) 8.24 0.95 13 Taste and flavor (C2) 60.25 6.92 5 Nutrition and health (C3) 31.51 3.62 9 Brand Image (B2) 5.62 4 Brand Awareness (C4) 7.68 0.43 14 Brand Reputation (C5) 18.63 1.05 12 Brand Innovation (C6) 73.70 4.14 7 Marketing and Promotion (B3) 54.80 1 Social Media Popularity (C7) 25.94 14.22 3 Endorsement or Recommendation by Internet Influencers (C8) 55.77 30.56 1 Price (C9) 11.24 6.16 6 Effectiveness of Marketing Activities (C10) 7.05 3.87 8 Consumption Experience (B4) 28.08 2 Purchase Convenience (C11) 25.15 7.06 4 Packaging Design (C12) 5.59 1.57 11 Cultural Experience (C13) 60.11 16.88 2 After-sales Service (C14) 9.15 2.57 10 3.3. Model Test Considering regional characteristics, experts’ suggestions and literature reports [16-18], Spicy Sticks (Liangpin Puzi), Duck Neck (Zhou black duck), Boneless Chicken Feet (Three Squirrels) and Luosifen (Li Ziqi) were selected to test the ICF consumer recognition model. Their latest reviews of network crawling on the JD.COM platform were analyzed based on the 14 indices of the evaluation system. The obtained raw indices data were subjected to dimensionless processing to facilitate subsequent data handling as shown in Table 17. The obtained scores of 4 kinds of ICF based on the comprehensive weight values calculation of the evaluation criteria, were shown in the Figure 2. The results of Luosifen (Li ziqi), Boneless chicken feet(Three squirrels), Duck neck(Zhou black duck)and Spicy bar (Liangpin puzi) were 0.290, 0.200, 0.280, 0.230 respectively. The difference in scores among the four popular foods on the internet was not significant, indicating that they were highly recognized by consumers as the popular products on the JD.COM platform. 3.4. Model Strengths and Sensitivities Since the weights obtained from AHP analysis depend on the importance levels of different projects manually specified, different input levels of importance by the experts will lead to different results. In order to conduct sensitivity analysis of the model, new judgment matrices were constructed based on the changes of the relative importance between each topic [19,20]. The importance value of Marketing and Promotion compared to Brand Image changed from 7 to 9, and judgment matrix A in Table 18 was constructed. On the basis of the above change, the importance value of Consumption Experience compared to Product Quality changed from 4 to 2, and judgment matrix B in Table 19 was constructed. 6 Table 17. Non-dimensionalized data for each indices ICF Luosifen(Li ziqi) Boneless chicken feet(Three squirrels) Duck neck (Zhou black duck) Spicy bar (Liangpin puzi) Ingredients and Raw Materials (C1) 0.07 0.21 0.46 0.27 Taste and Flavor (C2) 0.31 0.23 0.27 0.18 Nutrition and Health (C3) 0.16 0.08 0.31 0.45 Brand Awareness (C4) 0.33 0.21 0.28 0.18 Brand Reputation (C5) 0.30 0.23 0.32 0.14 Brand Innovation (C6) 0.30 0.09 0.43 0.17 Social Media Popularity (C7) 0.27 0.24 0.31 0.17 Endorsement or Recommendation by Internet Influencers (C8) 0.27 0.25 0.24 0.24 Price (C9) 0.29 0.21 0.30 0.20 Effectiveness of Marketing Activities(C10) 0.36 0.17 0.27 0.20 Purchase Convenience (C11) 0.25 0.20 0.32 0.23 Packaging Design (C12) 0.22 0.20 0.36 0.22 Cultural Experience (C13) 0.37 0.13 0.22 0.28 After-sales Service (C14) 0.25 0.21 0.33 0.21 Figure 2. Scores of the internet celebrity food According to the two judgment matrices (A) and (B), the weights of the evaluation criteria obtained by the AHP method were shown in Table 20 and 21 respectively, which have passed the consistency test. The obtained weights values were used to evaluate the test samples, and the scoring situations were shown in Table 22. It can be seen that although the numerical values of the scores of four testing ICF varied under these two changes, their relative sorting maintained consistent with the situation before the change. Luosifen(Li ziqi) ranked first and the others were in order, which indicated that the stability of the model was relatively good [21,22]. Table 18. Topic level judgment matrix (A) Topics Product Quality (B1) Brand Image (B2) Marketing and Promotion (B3) Consumption Experience (B4) Product Quality (B1) 1 3 0.2 0.25 Brand Image (B2) 0.333 1 0.111 0.2 Marketing and Promotion (B3) 5 9 1 3 Consumption Experience (B4) 4 5 0.333 1 Table 19. Topic level judgment matrix (B) Topics Product quality (B1) Brand image (B2) Marketing and promotion (B3) Consumption experience (B4) Product quality (B1) 1 3 0.2 0.5 Brand image (B2) 0.333 1 0.111 0.2 Marketing and promotion (B3) 5 9 1 3 Consumption experience (B4) 2 5 0.333 1 4. Discussion Based on the Analytic Hierarchy Process (AHP), weight values of 14 indices were calculated to establish an evaluation model for consumer recognition of ICF. This facilitate the 7 identification of the role of various influencing factors on consumer recognition of ICF. In summary, the reason why ICF became a hot topic of discussion was not only because of their delicious taste and uniqueness but also because of the cultural connotation and innovative spirit they embodied. In this rapidly changing era, only by constantly pursuing innovation and excellence could one have ridden the waves and led the trend in the ocean of cuisine. Table 20. Results of AHP Hierarchical Analysis (A) Topics Characteristic Vectors Weight Values (%) Largest Eigenvalue CI Product Quality (B1) 0.441 11.031 4.130 0.043 Brand Image (B2) 0.200 5.008 Marketing and Promotion (B3) 2.266 56.653 Consumption Experience (B4) 1.092 27.307 Table 21. Results of AHP Hierarchical Analysis (B) Topics Characteristic Vectors Weight Values (%) Largest Eigenvalue CI Product Quality (B1) 0.515 12.867 4.037 0.012 Brand Image (B2) 0.206 5.142 Marketing and Promotion (B3) 2.346 58.660 Consumption Experience (B4) 0.933 23.331 Table 22. Scores of the testing internet celebrity food Number SDE Score(A) Score(B) 1 Luosifen (Li ziqi) 0.287 0.285 2 Duck neck (Zhou black duck) 0.278 0.279 3 Spicy bar (Liangpin puzi) 0.230 0.229 4 Boneless chicken feet (Three squirrels) 0.205 0.207 Upon comparison of 4 test samples from online shopping platforms, the scores of the Luosifen (Li ziqi) was the highest, while the scores of the Boneless chicken feet (Three squirrels) was lowest. Luosifen (Li ziqi) performed relatively excellently in the evaluation of the established model. Endorsement or Recommendation by Internet Influencers and Cultural Experience may be the reasons for its high score. Li Ziqi, as the "top tier" among internet celebrities, has a high level of influence in the online public opinion arena. In the fast-paced modern life, Luosifen (Li Ziqi) promotional videos present a poetic sense of slow life, satisfying people who have lived in the city for a long time to imagine another way of life, have beautiful memories and expectations of local culture, and yearn to return to nature, and temporarily escape from the anxiety of reality. The high recognition of Luosifen (Li Ziqi) by consumers indicated that cultural connotation was increasingly becoming the key to the success of ICF, which was consistent with the previous research [23]. This established model passed the testing of strengths and sensitivities, representing the good stability. It could make quantitatively informed decisions about which ICF best fits the criteria of consumer recognition. However, due to the strong subjectivity of the AHP method, the universality of the established model was not strong. Therefore, in future work, more test samples will be selected to verify and improve the effectiveness of the model. 5. Conclusion In this study, an evaluation indices system for consumer recognition of ICF containing 14 evaluation indices have been established. Based on this indices system, 5 judgment matrixes were constructed by the experts. Through the calculation of AHP method, the comprehensive weight value of Endorsement or Recommendation by Internet Influencers, Cultural Experience, and Social Media Popularity were above 14%, which plays a very important role in the evaluation of ICF. Luosifen (Li ziqi), Boneless chicken feet (Three squirrels), Duck neck(Zhou black duck)and Spicy bar (Liangpin puzi), were selected to test the established model. The scores of the Luosifen(Li ziqi) were the highest due to its Endorsement or Recommendation by Internet Influencers and Cultural Experience. On the basis of changing the judgment matrix, a sensitivity analysis was conducted on the established model, and the results indicate that the model had relatively good stability. In summary, these research findings will provide theoretical guidance for the development of ICF. References [1] X. Li, Y. Wang, “Marketing strategies of Internet-famous food,” Food Industry 2020, 41(5), 123- 126. [2] H. Zhang, M. Zhao, “Analysis of the popularity mechanism of Internet-famous food,” Modern Business Trade Industry 2021, 42(3), 56-58. [3] Q. W. Cao, “The consumption logic of E-commerce food from the perspective of internet modernity,” E-Commerce Letters 2024, 13(2), 1725-1730. [4] L. N. Wang, “Innovation of Visual Communication Design of Interactive Packaging for Internet-Famous Food Based on Artificial Intelligence,” Scientific Programming 2022, 15, 5828852.1-5828852.7. [5] C. Song, L. Zheng, X. J. Shan, “An analysis of public opinions regarding Internet-famous food: a 2016–2019 case study on Dianping,” British Food Journal 2022, 124 (12), 4462. [6] L. Q. Huang, Y. Yan, W. S. Niu, “Exploring the influence fFactors of repeated purchase intention in the context of online celebrity,” China Journal of Information Systems 2025, 29, 1- 19. [7] W. Fei, Z. X. An, “Analysis of the formation mechanism of consumer trust under the background of internet celebrity food online shopping,” Economy and Managenment 2023, 37, 29- 37. [8] N. Polakis, N. Koupatou, A. Tsouchlaraki, “Evacuating populations with special needs using analytical hierarchy process (AHP) method and GIS,” Journal of Geographic Information System 2024, 16(4), 276-288. [9] A. Rostami, E. Hamzehzadeh, A. F. Mehrabad, et al. “Analytic hierarchy process (AHP) of factors influencing on sport participation of high school students,” Advances in Environmental Biology 2014, 8(22), 855-860. [10] T. L. Saaty, “Decision making with the analytic hierarchy process,” International Journal of Services Sciences 2008, 1(1), 83-98. 8 [11] T. L. Saaty, “Correction to: Some mathematical concepts of the analytic hierarchy process,” Behaviormetrika 2021, 48, 193- 194. [12] Z. Y. Zhang, “Acquisition of metadata efficiently by using cctopus web crawler technology in university libraries,” Modern Information Technology 2019, 3(04), 4-6. [13] D. F. Cui, C. J. Liu, “Study on the influence of internet celebrities food experience and consumers' online purchase intention,” Xinjiang State Farms Economy 2024, 5, 1-19. [14] F. Liu, L. Meng, S. Chen, “The impact of network celebrities’ information source characteristics on purchase intention,” Chinese Journal of Management 2020,17(1), 94-104. [15] M. P. Krishna, K. Himanshu, et al. “Influencer marketing: A new frontier for brand awareness in the digital era,” International Journal for Research in Applied Science & Engineering Technolo. 2023, 11(10), 1752-1773. [16] X. G. Lei, Z. Huang, C. Wu, et al. “A comparison of luosifen Study and pasta study: A bibliometric analysis,” Food Culture & Society 2023, 26(2), 210-225. [17] Y. W. Chen, “Product innovation and consumer preference in leisure snack industry: Evidence from zhou hei ya,” Journal of Food Products Marketing 2023, 29(3), 112-128. [18] Y. Wu, X. Cai, “Analysis of the current situation of snack enterprises in the post-epidemic era-taking bestore as an example,” Finance 2023, 13(6), 1–3. [19] I. Vinogradova-Zinkevič, “Comparative sensitivity analysis of some fuzzy AHP methods,” Mathematics 2023, 11(24), 4984. [20] C. Acar, E. Haktamir, G. T. Temur, et al. “Sustainable stationary hydrogen storage application selection with interval- valued intuitionistic fuzzy AHP,” Sustainable Energy Technologies and Assessments 2023, 58, 103-118. [21] N. M. M. Torre, V. A. P. Salomon, A. K. Florek-Paszkowska, “Multi-criteria classification of spare parts in the steel industry,” Case Studies in Engineering 2025, 8(2), 102-115. [22] E. J. Yazo-Cabuya, J. A. Herrera-Cuartas, A. Ibeas, “Organizational risk prioritization using DEMATEL and AHP towards Sustainability,” Sustainability 2024, 16, 1080. [23] G. Xu, N. Suwandej, “The relationship among brand culture, marketing strategy, technical innovation and customer engagement behavior toward liuzhou snail noodle: Mediating of service quality and trust,” Journal of Business Research 2023, 166, 114-125.