Risk analysis of social media content based on neural network classification of a message text emotional coloring

One of the promising areas of Data Science within the framework of practice-oriented approaches to the analysis of social networks (Social network analysis) from the point of view of network users’ (agents’) opinion formalization is a class of content analysis methods designed for automated identification of emotionally colored vocabulary in texts and emotional evaluation of authors in relation to the objects referred to in the text. With the help of such an analysis, it is possible to study an array of messages and other data and determine how they are emotionally colored - positively, negatively or neutrally. The article offers a comparative analysis of two approaches to the study of text sequences classification possibilities depending on their emotional coloring: one by means of a recurrent neural network (RNN) and another involving graph convolutional networks (GCN). The first approach is implemented through deep learning utilizing the Deep Learning Designer tool (MathWorks © MATLAB R2021b). The second approach is based on the application of convolutional graph neural networks for text classification. GCN implementation is carried out in Python using the appropriate set of libraries for data analysis. In addition, the paper shows that the resulting model can be used in risk assessment, where the resulting value serves as a correction factor in calculating the risk of user involvement. Based on the results of the two approaches comparison, it is shown that when using GCN, the percentage of training data decreases, which indicates the sensitivity of the method to a smaller amount of training data, while the accuracy of the model increases with comparable configurable training parameters

1. Aggarwal, C. C., Zhai, C. X. A survey of text classification algorithms. In Mining text data. Springer. In Mining Text Data. 2012;163-22.

2. Ostapenko A.G., Chapurin E.Yu., Kalashnikov A.O., Ostapenko O.A., Ostapenko G.A. social media and risk monitoring. Ed. Corresponding Member RAS D.A. Novikov. M.: Hotline-Telecom; 2019;(4). (In Russ.)

3. Hochreiter, S., and Schmidhuber, J. Long short-term memory. Neural computation. 1997.9(8):1735-1780.

4. Battaglia, P.W.; Hamrick, J.B.; Bapst, V.; Sanchez-Gonzalez, A.; Zambaldi, V.; Malinowski, M.; Tacchetti, A.; Raposo, D.; Santoro, A.; Faulk-ner, R.; et al. Relational inductive biases, deep learning, and graph networks. arXiv preprint. 2018;(1).

5. Dataset sentiment analysis with tweets. Available at: https://www.kaggle.com/vandalko/keras-lstm-twitter-sentiment-analysis/data

6. Twitter API. Available at: https://developer.twitter.com/en/products/twitter-api

7. Deep learning for humans. Available at: https://keras.io/

8. Yao, Liang, Chengsheng Mao, and Yuan Luo. "Graph convolutional networks for text classification. Proceedings of the AAAI Conference on Artifi-cial Intelligence. 2019;33: 7370-7376.

9. Rafael E. Banchs. Text Mining with MATLAB. Springer International Publishing; 2012. 468 p.

10. Accuracy and Loss. Available at: https://docs.paperspace.com/machine-learning/wiki/accuracy-and-loss

11. Chapurin E.Yu. Toolkit for the study of distributed computer systems in the context of the spread of viral content: thematic modeling of malware. Informatsiya i bezopasnost' = Information and security. 2020;23 (2):291-304. (In Russ.)

12. Sentiment Analysis template. Available at: https://github.com/floydhub/sentiment-analysis-template

13. Belonozhkin V.I., Dergachev Yu.A., Turchin A.S. Methodology for assessing and regulating risks in the operation of software tools that form a technical channel of information leakage due to program-controlled incidental electromagnetic radiation Informatsiya i bezopasnost' = Information and security. 2020;23(1):51-66. (In Russ.)

14. Ruzhitskiy E., Schwarzkopf E.A., Manmareva V.V. Risk ranking of publicly available Internet resources based on average daily measurements of information processes of perception of their content by users. Informatsiya i bezopasnost' = Information and security. 2020;23(1):97-106. (In Russ.)

15. Ermakov S.A., Katsenko Ya.M., Bolgov A.A. Assessment and regulation of risks of violation of information security of telecommunication communication networks and industrial Internet of things control. Informatsiya i bezopasnost' = Information and security. 2020;23(1):107-114. (In Russ.)

16. Ostapenko A.G. Infodemia and Social Networks: Induced Risks and Opportunities. Informatsiya i bezopasnost' = Information and security. 2020;23(2):235-244. (In Russ.)

17. Ostapenko A.G. «Infodemia» and social networks: models of the epidemic process. Informatsiya i bezopasnost' = Information and security. 2020;23(2):285-290. (In Russ.)

18. Ostapenko A.G., Sokolova E.S., Pasternak Yu.G. Formalization of the description of monoviral epidemic processes in networks. Informatsiya i bezopasnost' = Information and security. 2020;23(4):497-510. (In Russ.)

19. Shtefanovich Y., Schwarzkopf E.A., Manmareva V.V. Vector assessment of the danger of spreading viral content based on the average daily reactivity of Internet users. Informatsiya i bezopasnost' = Information and Security. 2020;23(1):79-86. (In Russ.)

20. Chapurin E.Yu., Guslyannikov A.E., Parinova L.V. A software and hardware complex for risk analysis of the descriptive content of social networks: structure, appearance and databases. Informatsiya i bezopasnost' = Information and security. 2020;23(3):389-398. (In Russ.)

21. Chapurin E.Yu., Guslyannikov A.E., Parinova L.V. Software and hardware complex for risk analysis of destructive content of social networks: main components and vulnerabilities. Informatsiya i bezopasnost' = Information and security. 2020;23(3):409-418. (In Russ.)

22. Ostapenko A.G., Ostapenko A.A., Lantyukhov N.M. On the issue of trends and tools of socio-informational global confrontation. Informatsiya i bezopasnost' = Information and security. 2020;23(4):519-524

23. Moskaleva E.A., Barannikov N.I., Karebin D.S. Cartographic study of the activities of cybercriminal groups in the context of increasing the effectiveness of protection measures. Informatsiya i bezopasnost' = Information and security. 2020;23(3):431-446. (In Russ.)

24. Heart A.L., Markov R.V., Gerasimov I.V. A cartographic approach to the study of the processes of dissemination of destructive content in communities of a single subject of the social network "VKontakte". Informatsiya i bezopasnost' = Information and security. 2020;23(2):203-214. (In Russ.)

25. Chapurin E.Yu., Guslyannikov AE, Razinkin KA Content destructiveness, its classifiers and scanners for risk analysis of social networks. Informatsiya i bezopasnost' = Information and security. 2020;23(3):375-378. (In Russ.)

26. Ruzhitskiy E., Schwarzkopf E.A., Manmareva V.V. Thematic classification of Internet resources based on vector illustration into groups of potentially dangerous content. Informatsiya i bezopasnost' = Information and security. 2020;23(1):123-132. (In Russ.)

27. Grechishkin A.V., Rakhmanin D.N., Sviridov A.V. Telecommunication and control modems: protection against malicious code injection attacks based on expert assessment of modem protection and risk management. Informatsiya i bezopasnost' = Information and security. 2020;23(2):305-314. (In Russ.)