Communicative Monitoring of Vaccination Prevention Based on the Analysis of Associative Series: Experience with the Application of Natural Language Processing Technologies

Relevance. Despite the proven effectiveness and cost-efficiency of immunoprophylaxis as a cornerstone of public health, its perception has become increasingly heterogeneous in contemporary settings, with vaccine hesitancy now recognized as a major global health threat. Conventional questionnaire-based methods for assessing attitudes toward vaccination often capture socially desirable responses, thereby limiting the detection of underlying cognitive patterns. Aim. This study aimed to evaluate the potential of natural language processing (NLP) techniques applied to associative word chains – elicited through open-ended interviews – as an auxiliary tool for communicative monitoring in the field of vaccination. Materials and Methods. The research involved 90 human respondents (45 adolescents aged 12–17 years and 45 adults aged 20–68 years), recruited via snowball sampling, and 532 virtual respondents generated by the Qwen2-7B-Instruct large language model. Participants were asked to provide up to 10 free associations to 14 stimulus words (12 vaccine-preventable diseases from the Russian National Immunization Schedule, plus the terms «vaccination» and «inoculation»), without any quantitative or qualitative restrictions on lexical choices. In total, we analyzed 65 320 associations. We applied lemmatization, clustering, and sentiment analysis using the «cointegrated/rubert-tiny-sentimentbalanced» model. Statistical analyses included the Mann–Whitney U test, Pearson’s chi-squared test, and odds ratio calculations (p <0.05). Data processing was performed using StatTech 4.10.3 and Python 3.11. Results and Discussion. Human respondents rarely completed full 10-word chains (47.3 % completion rate), whereas the AI consistently generated all 10 responses – though some contained unreadable symbols and were excluded. From all responses, we compiled a lexicon of 4 096 unique words and phrases; lexical overlap between adolescents and adults reached 23.6 % (30.7 % among AI-generated responses). Positive and negative associations appeared with roughly equal frequency overall, and semantic field alignment occurred in 28.1 % of cases. All associations clustered into four thematic categories: «Disease Characteristics», «Vaccination», «Danger», and «Other». Regardless of age or respondent type, the first association typically reflected «Disease Characteristics», while references to vaccination emerged mainly from the second or third response onward. Adults produced significantly more vaccination-related associations than adolescents. Moreover, the presence of at least one term from the «Vaccination» cluster increased the odds of a high selfreported attitude toward immunization (on a 10-point scale) by 4.3 times (p = 0.021). Sentiment analysis revealed an overall ratio of 1.68 positive to every negative association; however, human respondents exhibited stronger sentiment polarization, whereas AI responses leaned toward neutrality — likely due to media-driven constraints on publishing negative content about vaccines. Although participants generally reported highly favorable attitudes toward vaccination (median = 9 [IQR: 7–10]), sentiment analysis of their associative responses indicated only a moderate level of acceptance. Conclusion. The findings demonstrate that the chain associative experiment, when combined with NLP methods, can effectively complement traditional tools of communicative monitoring by uncovering latent cognitive structures and offering a more nuanced assessment of vaccination acceptance. Consequently, integrating this approach into epidemiological surveillance systems appears both feasible and methodologically justified.

1. Steinert J.I., Sternberg H., Prince H., et al. COVID-19 vaccine hesitancy in eight European countries: Prevalence, determinants, and heterogeneity // Sci. Adv. 2022. Vol. 8, № 17. P. eabm9825. DOI: 10.1126/sciadv.abm9825.

2. Brumbaugh K.Q., Gellert F., Mokdad A.H. Understanding Vaccine Hesitancy: Insights and Improvement Strategies Drawn from a Multi-Study Review // Vaccines (Basel). 2025. Vol. 13, № 10. P. 1003. DOI: 10.3390/vaccines13101003.

3. World Health Organization. Ten threats to global health in 2019 [Electronic resource]. URL: https://www.who.int/ru/news-room/spotlight/ten-threats-to-global-healthin-2019 (accessed: 20.01.2026).

4. Nehal K.R., Steendam L.M., Campos Ponce M., et al. Worldwide Vaccination Willingness for COVID-19: A Systematic Review and Meta-Analysis // Vaccines (Basel). 2021. Vol. 9, № 10. P. 1071. DOI: 10.3390/vaccines9101071.

5. Kuznetsova O.Yu., Khalimov Yu.Sh., Moiseeva I.E., et al. Priverzhennost’ vaktsinoprofilaktike i uroven’ znanii: est’ li sviaz’? // Rossiiskii semeinyi vrach. 2023. Vol. 27, № 2. P. 47–53. (In Russ.). DOI: 10.17816/RFD501781.

6. Minaeva V.A., Golubkova A.A. Otnoshenie k vaktsinoprofilaktike sotrudnikov detskoi polikliniki // Infektsionnye bolezni: Novosti. Mneniia. Obuchenie. 2025. Vol. 14, № 1 (52). P. 64–73. (In Russ.)

7. Briko N.I., Mindlina A.Ya., Galina N.P., et al. Priverzhennost’ razlichnykh grupp naseleniia immunoprofilaktike: kak izmenit’ situatsiiu? // Fundamental’naia i klinicheskaia meditsina. 2019. Vol. 4, № 4. P. 8–18. (In Russ.). DOI: 10.23946/2500-0764-2019-4-4-8-18.

8. Goroshko N.V., Emel’ianova E.K., Patsala S.V. Problema kovid-antivaktsinatorstva: Rossiia na mirovom fone // Sotsial’nye aspekty zdorov’ia naseleniia. 2021. Vol. 67, № 4. (In Russ.)

9. Lanz L., Thielmann I., Gerpott F.H. Are social desirability scales desirable? A meta-analytic test of the validity of social desirability scales in the context of prosocial behavior // J. Pers. 2022. Vol. 90, № 2. P. 203–221. DOI: 10.1111/jopy.12662.

10. Bispo Júnior J.P. Social desirability bias in qualitative health research // Rev. Saude Publica. 2022. Vol. 56. P. 101. DOI: 10.11606/s1518-8787.2022056004164.

11. Briko N.I., Fel’dblium I.V., Alyeva M.Kh., et al. Kontseptsiia risk-kommunikatsii po obespecheniiu priverzhennosti k vaktsinatsii kak neobkhodimaia sostavliaiushchaia strategicheskogo razvitiia immunoprofilaktiki v Rossii // Obshchestvennoe zdorov’e. 2021. Vol. 1, № 1. P. 32–43. (In Russ.)

12. Plakida A.V., Briko N.I., Namazova-Baranova L.S., et al. Povyshenie priverzhennosti naseleniia vaktsinatsii: otsenka i sistemnyi podkhod k realizatsii // Epidemiologiia i vaktsinoprofilaktika. 2022. Vol. 21, № 3. P. 4–26. (In Russ.). DOI: 10.31631/2073-3046-2022-21-3-4-26.

13. Bavrina A.P. Osnovnye poniatiia statistiki // Meditsinskii al’manakh. 2020. № 3 (64). P. 101–111. (In Russ.)

14. Primakov V.L. K voprosu o klassifikatsii metodov formirovaniia vyborok v sotsiologicheskikh issledovaniiakh // Vestnik Moskovskogo gosudarstvennogo lingvisticheskogo universiteta. Obshchestvennye nauki. 2024. № 3 (856). P. 114–121. (In Russ.)

15. Borisova Yu.A. Assotsiativnyi eksperiment v sovremennykh psikholingvisticheskikh issledovaniiakh // Psikhologiia. Istoriko-kriticheskie obzory i sovremennye issledovaniia. 2019. Vol. 8, № 1–1. P. 265–275. (In Russ.)

16. Qwen2-7B-Instruct [Electronic resource]. Hugging Face, 2025. URL: https://huggingface.co/Qwen/Qwen2-7B-Instruct (accessed: 20.01.2026).

17. Toporkov O., Agerrí R. On the Role of Morphological Information for Contextual Lemmatization // Computational Linguistics. Cambridge, MA: MIT Press, 2024. Vol. 50, № 1. P. 157–191. DOI: 10.1162/coli_a_00497.

18. rubert-tiny-sentiment-balanced [Electronic resource]. Hugging Face. URL: https://huggingface.co/cointegrated/rubert-tiny-sentiment-balanced (accessed: 20.01.2026).

19. Criminal Code of the Russian Federation dated 13.06.1996 № 63-FZ (latest revision) [Electronic resource]. ConsultantPlus. URL: https://www.consultant.ru/document/cons_doc_LAW_10699/ (accessed: 20.01.2026). (In Russ.)

20. I ttefaq M., Vu H.T., Zain A., et al. Analysis of public opinion polls about COVID-19 vaccines: Theoretical and policy implications for vaccine communication and campaigns to address vaccine hesitancy // Hum. Vaccin. Immunother. 2024. Vol. 20, № 1. P. 2437921. DOI: 10.1080/21645515.2024.2437921.

21. Lazarus J.V., Wyka K., White T.M., et al. A survey of COVID-19 vaccine acceptance across 23 countries in 2022 // Nat. Med. 2023. Vol. 29. P. 366–375. DOI: 10.1038/s41591-022-02185-4.

22. Romanenko V.V., Kosova A.A., Smirnova S.S., et al. Organizatsiia sistemy immunoprofilaktiki infektsionnykh boleznei v Sverdlovskoi oblasti: uchebnoe posobie. Ekaterinburg: USMU, 2023. 172 p. ISBN 978-5-00168-053-6. (In Russ.)