The paper presents key approaches to understanding and researching radicalisation, as well as the opportunities and limitations of applying some research methods to model network topology and assess content similarity of online communities. Today, Web Mining and AI methods and technologies are often applied in research on social networks and youth participation in them. However, the question how these approaches can be effectively used to study online radicalisation remains open. The answer to this question should increase the explanatory and predictive power of computational models for detecting and predicting radicalisation in the online space. In much of the Russian research on online radicalisation, a common approach has been to reduce the task of identifying the interconnectedness of individual online communities or clusters of them to assessing the degree of similarity in terms of subscribers or linguistic markers. This approach is limited in predicting new connections between communities and justifying radicalisation pathways, but is relevant in modelling information diffusion. In this paper, the authors aim to demonstrate the possibilities and limitations of applying the tf-idf, doc2vec methods to assess the content similarity of online communities without signs of radicalisation and online communities with signs of radicalisation. This approach allowed the authors to identify communities with a significant tendency to unite (to establish direct links). The paper presents the results of the comparative study in the form of social graphs formed according to the principles of subscriber commonality, similarity of significant words, and contextual similarity based on the doc2vec model. The social graph based on doc2vec method performed better in terms of clustering of online communities as well as interpretability of the results. This is crucial for detecting and predicting radicalisation online, as it opens the prospect of exploring the nature of assortativity in the observed network.

social network, community, radicalisation, network topology, tf-idf, doc2vec

1. Akhremenko A. S., Stukal D. K., Petrov A. P. Network or text? Factors of protest dissemination in social media: theory and data analysis. Polis. Politicheskie issledovaniya, 2020: 2: 73–91 (in Russ.). DOI: 10.17976/jpps/2020.02.06; EDN: APZWMB.
2. Karpova A. Y., Kuznetsov S. A. et al. Method for searching images with signs of ultra-right radicalization in social media based on neural network classification. Sistemy upravleniya i informacionny`e texnologii, 2023: 1(91): 59–64 (in Russ.). DOI: 10.36622/VSTU.2023.91.1.012; EDN: KGWMLC.
3. Karpova A. Yu., Savelev A. O. Opportunities and boundaries of Big Data technologies for the study of online radicalization. In Third December socio-political readings "How you live, Russia?". Challenges of the pandemic, parliamentary elections and strategic agenda for society and the state. Moscow, FNISC RAN, 2022: 67–78 (in Russ.). EDN: ZXFWGV.
4. Karpova A. Y., Savelev A. O. et al. Ultra-right radicalization: methodology of automated threat detection by web mining methods. Vestnik RFFI. Gumanitarnye i obshhestvennye nauki, 2020: 5(102): 30–43 (in Russ.). DOI: 10.22204/2587-8956-2020-102-05-30-43; EDN: BEZTTC.
5. Karpova A. Y., Savelev A. O. et al. Studying Online Radicalization of Youth through Social Media (Interdisciplinary Approach). Monitoring obschestvennogo mneniya: ekonomicheskie i social’nye peremeny, 2020: 3(157): 159–181 (in Russ.). DOI: 10.14515/monitoring.2020.3.1585; EDN: CXJJUW.
6. Karpova A. Y., Savelev A. O., Kuznetsov S. A. Transformation of social practices of responsible fatherhood into deviant forms of social activity. Vektory blagopoluchiya: ekonomika i socium, 2021: 4(43): 107–118 (in Russ.). DOI: 10.18799/26584956/2021/4(43)/1130; EDN: DDROPU.
7. Karpova A. Yu., Shirykalov A. M. Development of a method for evaluating the similarity of text collections using their vector representations obtained by the doc2vec method. In Youth and Modern Information Technologies. Tomsk, TPU, 2021: 75–76 (in Russ.). EDN: SWSZVQ.
8. Kuznetsov S. A., Karpova A. Yu., Savelev A. O. Methods and technologies of intellectualization of search for destructive and radical content in social media: analysis of the current state. Vestnik kompyuternyh i informacionnyh texnologij, 2023: 4(226): 39–48 (in Russ.). DOI: 10.14489/vkit.2023.04.pp.039-048; EDN: EVCQQC.
9. Savelev A. O., Karpova A. Yu. A. Approach to the assessment of interconnections of social media communities on the basis of thematic similarity of textual content (on the example of communities of socially active fathers and communities with signs of radicalization). Kazan Economic Bulletin, 2022: 1(57): 96–103 (in Russ.). EDN: DNZYVS.
10. Sokolova T. V., Chepovsky A. M. The task of analyzing social network user profiles. In Situation centers and information-analytical systems of class 4i for monitoring and security tasks (SCVRT2017). Moscow–Protvino, IFTI, 2017: 198–201 (in Russ.). EDN: YSHYDJ.
11. Stukal D. K., Akhremenko A. S., Petrov A.P. Affective political polarization and hate speech: created for each other? Vestnik RUDN. Ser.: Politologiya, 2022: 3: 480–498 (in Russ.). DOI: 10.22363/2313-1438-2022-24-3-480-498; EDN: VLTQRN.
12. Chuprov V. I., Zubok Yu. A. Molodezhnyj ekstremizm: sushhnost, formy proyavleniya, tendencii [Youth extremism: essence, forms of manifestation, trends]. Moscow, Academia, 2009: 320 (in Russ.). EDN: SUFKBZ.
13. Akram M., Nasar A. Systematic Review of Radicalization through Social Media. Ege Academic Review. 2023. Vol. 23(2). P. 279–296. DOI: 10.21121/eab.1166627; EDN: JFWLSR.
14. Alvari H., Sarkar S., Shakarian P. Detection of Violent Extremists in Social Media. In 2nd International Conference on Data Intelligence and Security (ICDIS). South Padre Island, TX, USA, 2019: 43–47. DOI: 10.1109/ICDIS.2019.00014.
15. Araque O., Iglesias C. A. An Approach for Radicalization Detection Based on Emotion Signals and Semantic Similarity. IEEE Access, 2020: 8: 17877–17891. DOI: 10.1109/access.2020.2967219.
16. Binder J. F., Kenyon J. Terrorism and the internet: How dangerous is online radicalization? Frontiers in Psychology, 2022: 13. DOI: 10.3389/fpsyg.2022.997390; EDN: RGSMYY.
17. Borum R. Radicalization into Violent Extremism I: A Review of Social Science Theories. Journal of Strategic Security, 2012: 4(4): 7–36. DOI: 10.5038/1944-0472.4.4.1.
18. Borum R. Radicalization into Violent Extremism II: A Review of Conceptual Models and Empirical Research. Journal of Strategic Security. 2012: 4(4): 37–62. DOI: 10.5038/1944-0472.4.4.
19. Borum R. Rethinking radicalization. Journal of Strategic Security, 2012: 4(4): 1–6. Accessed 15.12.2023. URL: https://digitalcommons.usf.edu/jss/vol4/iss4/1
20. Borum R. Understanding the terrorist mind-set. FBI Law Enforcement Bulletin, 2003: 72(7): 7–10.
21. Clancy T., Addison B. et al Contingencies of Violent Radicalization: The Terror Contagion Simulation. Systems, 2021: 9(4): 90. DOI: 10.3390/systems9040090.
22. Das S., Biswas A. The Ties that matter: From the perspective of Similarity Measure in Online Social Networks. ArXiv, 2022. DOI: 10.48550/arXiv.2212.10960.
23. de la Roche R. S. Why is collective violence collective? Sociological Theory, 2001: 19(2): 126–144. DOI: 10.1111/0735-2751.00133.
24. Deem A. The Digital Traces of #whitegenocide and Alt-Right Affective Economies of Transgression. International Journal of Communication, 2019: 13: 3183–3202.
25. Derbas N., Dusserre E. et al. Eventfully Safapp: hybrid approach to event detection for social media mining. In Journal of Ambient Intelligence and Humanized Computing, 2020: 11(1): 87–95. DOI: 10.1007/s12652-018-1078-7; EDN: VCUFZA.
26. Ducol B. A Radical sociability: in defense of an online/offline multidimensional approach to radicalization. Social Networks, Terrorism and Counter-Terrorism: Radical and Connected. Ed by M. Bouchard. New York, Routledge, 2015: 82–104. 
27. Expressions of Radicalization: Global Politics, Processes and Practices. Cambridge, Palgrave Macmillan, 2018: 526. DOI: 10.1007/978-3-319-65566-6.
28. Ferrara E. Contagion dynamics of extremist propaganda in social networks. Information Sciences, 2017: 418–419: 1–12. DOI: 10.1016/j.ins.2017.07.030.
29. Floridi L. Hyperhistory and the philosophies of information policies. The Onlife Manifesto, 2015: 51–63. DOI: 10.1007/978-3-319-04093-6_12.
30. Francisco M., Castro J. L. A fuzzy model to enhance user profiles in microblogging sites using deep relations. Fuzzy Sets and Systems, 2020: 401: 133–149. DOI: 10.1016/j.fss.2020.05.006.
31. Garcet S. Understanding the psychological aspects of the radicalization process: a sociocognitive approach. Forensic Sciences Research, 2021: 6(2): 115–123. DOI: 10.1080/20961790.2020.1869883.
32. Gezha V. N., Kozitsin I. V. The Effects of Individuals’ Opinion and Non-Opinion Characteristics on the Organization of Influence Networks in the Online Domain. Computers, 2023: 12(6). DOI: 10.3390/computers12060116.
33. Greenberg K. J. Counter-radicalization via the internet. The Annals of the American Academy of Political and Social Science, 2016: 668(1): 165–179. DOI: 10.1177/0002716216672635.
34. Grover T., Mark G. Detecting Potential Warning Behaviors of Ideological Radicalization in an Alt-Right Subreddit. In Proceedings of the International AAAI Conference on Web and Social Media, 2019: 13: 193–204. DOI: 10.1609/icwsm.v13i01.3221.
35. Hall M., Logan M. et al. Do Machines Replicate Humans? Toward a Unified Understanding of Radicalizing Content on the Open Social Web. Policy & Internet. 2020: 12: 109–138. DOI: 10.1002/poi3.223.
36. Hamm M., Spaaj R. Lone Wolf Terrorism in America: Using Knowledge of Radicalization Pathways to Forge Prevention Strategies. Washington DC: US Department of Justice, 2015. Accessed 15.12.2023. URL: https://www.ojp.gov/pdffiles1/nij/grants/248691.pdf
37. Horgan J. From Profiles to Pathways and Roots to Routes: Perspectives from Psychology on Radicalization into Terrorism. The Annals of the American Academy of Political and Social Science, 2008: 618(1): 80–94. DOI: 10.1177/0002716208317539; EDN: JLEUBT.
38. Ishfaq U., Khan H. U., Iqbal S. Identifying the influential nodes in complex social networks using centrality-based approach. Journal of King Saud University. Computer and Information Sciences, 2022: 34(10): 9376–9392. DOI: 10.1016/j.jksuci.2022.09.016; EDN: DKHGZY.
39. Karpova A. Yu., Kuznetsov S. A. et al. An online scan of extreme-right radicalization in social networks (the case of the Russian social network VKontakte). Journal of Siberian Federal University. Humanities and Social Sciences, 2022: 15: 12: 1738–1750. DOI: 10.17516/1997-1370-0948; EDN: IGVUQA.
40. Karpova A., Savelev A. et al. Method for detecting far-right extremist communities on social media. Social Sciences, 2022: 11: 5. DOI: 10.3390/socsci11050200; EDN: EFNFLP.
41. Kozitsin I. V. A general framework to link theory and empirics in opinion formation models. Scientific Reports, 2022: 12(1): 5543. DOI: 10.1038/s41598-022-09468-3; EDN: VVQEUT.
42. Kozitsin I. V. Opinion dynamics of online social network users: a micro-level analysis. The Journal of Mathematical Sociology, 2023: 47(1): 1–41. DOI: 10.1080/0022250X.2021.1956917; EDN: BHVFFM.
43. LaFree G. Lone-Offender Terrorists. Criminology and Public Policy, 2013: 12(1): 59–62. DOI: 10.1111/1745-9133.12018.
44. Lara-Cabrera R., Pardo A. G. et al. Measuring the Radicalisation Risk in Social Networks. IEEE Access, 2017: 5: 10892–10900. DOI: 10.1109/access.2017.2706018; EDN: GFVFJT.
45. Lee D.-H., Kim Y.-R. et al. Fake News Detection Using Deep Learning. Journal of Information Processing Systems, 2019: 15(5): 1119–1130. DOI: 10.3745/JIPS.04.0142.
46. McCauley C., Moskalenko S. Mechanisms of Political Radicalization: Pathways Toward Terrorism. Terrorism and Political Violence, 2008: 20(3): 415–433. DOI: 10.1080/09546550802073367.
47. Moghaddam F. M. The staircase to terrorism: A psychological exploration. American Psychologist, 2005: 60(2): 161–169. DOI: 10.1037/0003-066X.60.2.161.
48. Mussiraliyeva S., Bolatbek M. et al. On detecting online radicalization and extremism using natural language processing. In 21st International Arab Conference on Information Technology. Giza, 2020: 9300086. DOI: 10.1109/ACIT50332.2020.9300086; EDN JNULVZ.
49. Neumann P. R. The trouble with radicalization. International Affairs, 2013: 89(4): 873–893. DOI: 10.1111/1468-2346.12049.
50. Petrov A. Countering Fake News with Contagious Inoculation and Debunking: A Mathematical Model. 2022 15th International Conference Management of large-scale system development (MLSD). Moscow, 2022: 1–4. DOI: 10.1109/MLSD55143.2022.9933991.
51. Petrov A., Akhremenko A., Zheglov S. Dual Identity in Repressive Contexts: An Agent-Based Model of Protest Dynamics. Social Science Computer Review, 2023: 41(6): 2249–2273. DOI: 10.1177/08944393231159953; EDN: LPVEKT.
52. Petrov A., Proncheva O. Modeling propaganda battle: Decision-making, homophily, and echo chambers. Communications in Computer and Information Science, 2018: 930: 197–209. DOI: 10.1007/978-3-030-01204-5_19; EDN: WTYCBE.
53. Radicalization and Variations of Violence. New Theoretical Approaches and Original Case Studies. Ed. by D. Beck, J. Renner-Mugono. Springer, 2023: 212. DOI: 10.1007/978-3-031-27011-6.
54. Renström E. A., Bäck H., Knapton H. M. Exploring a pathway to radicalization: The effects of social exclusion and rejection sensitivity. Group Processes & Intergroup Relations, 2020: 23(8): 1204–1229. DOI: 10.1177/1368430220917215.
55. Rowe M., Saif H. Mining Pro-ISIS Radicalisation Signals from Social Media Users. In Proceedings of the Tenth International AAAI Conference on Web and Social Media (ICWSM 2016). 2016: 10(1): 329–338. DOI: 10.1609/icwsm.v10i1.14716.
56. Sharif W., Mumtaz S. et al. An empirical approach for extreme behavior identification through tweets using machine learning. Applied Sciences, 2019: 9(18). DOI: 10.3390/app9183723; EDN: VKLQVT.
57. Siebl T. Digital transformation: survive and thrive in an era of mass extinction. New York, Rosetta Books, 2019: 256.
58. Smith L. G., Wakeford L. et al. Detecting psychological change through mobilizing interactions and changes in extremist linguistic style. Computers in Human Behavior, 2020: 108: 106298. DOI: 10.1016/j.chb.2020.106298.
59. Tang L., Liu H. Community Detection and Mining in Social Media. Morgan & Claypool Publishers, 2010: 138. DOI: 10.1007/978-3-031-01900-5.
60. Tausch N., Bode S., Halperin E. Emotions in Violent Extremism. Handbook of the Psychology of Violent Extremism. Cambridge University Press, 2024.
61. The Routledge handbook of terrorism research. London, Routledge, 2011: 736.
62. Thompson R. Radicalization and the Use of Social Media. Journal of Strategic Security, 2011: 4: 167–190. DOI: 10.5038/1944-0472.4.4.8.
63. Tsapatsoulis N., Djouvas C. Opinion Mining From Social Media Short Texts: Does Collective Intelligence Beat Deep Learning? Frontiers in Robotics and AI, 2019: 5: 138. DOI: 10.3389/frobt.2018.00138.
64. Valentini D., Lorusso A. M., Stephan A. Onlife Extremism: Dynamic Integration of Digital and Physical Spaces in Radicalization. Frontiers in Psychology, 2020: 11: 524. DOI: 10.3389/fpsyg.2020.00524.
65. Wadhwa P., Bhatia M. P. S. An approach for dynamic identification of online radicalization in social networks. Cybernetics and Systems, 2015: 46(8): 641–665. DOI: 10.1080/01969722.2015.1058665.
66. Whittaker J. Rethinking Online Radicalization. Perspectives on Terrorism, 2022: 16(4): 27–40. Accessed 15.12.2023. URL: https://www.jstor.org/stable/27158150.
67. Winter Ch., Neumann P. et al. Online extremism: research trends in internet activism, radicalization, and counter-strategies. International Journal of Conflict and Violence, 2020: 14: 1–20. DOI: 10.4119/ijcv-3809.
68. Wojcieszak M. Carrying online participation offline: mobilization by radical online groups and politically dissimilar offline ties. Journal of Communication, 2009: 59(3): 564–586. DOI: 10.1111/j.1460-2466.2009.01436.x.
69. Xu G., Meng Y. et al. Sentiment Analysis of Comment Texts Based on BiLSTM. IEEE Access, 2019: 7: 51522–51532. DOI: 10.1109/ACCESS.2019.2909919.
70. Zafarani R., Abbasi M. A., Liu H. Social Media Mining: An Introduction. Cambridge University Press, 2014: 332. DOI: 10.1017/CBO9781139088510.
71. Zareie A., Sheikhahmadi A. et al. Finding influential nodes in social networks based on neighborhood correlation coefficient. Knowledge-Based Systems, 2020: 194: 105580. DOI: 10.1016/j.knosys.2020.105580.