2025, issue 1, p. 81-88
Received 30.01.2025; Revised 14.02.2025; Accepted 25.03.2025
Published 28.03.2025; First Online 30.03.2025
https://doi.org/10.34229/2707-451X.25.1.8
Previous | FULL TEXT (in Ukrainian) | Next
Using Machine Learning Methods to Develop a System of Social Dynamics
Violeta Tretynyk 1 * 
, Yulia Nad 2
1 The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute"
2 GXperts GmbH, Vienna
* Correspondence: This email address is being protected from spambots. You need JavaScript enabled to view it.
Introduction. The modern world is characterized by rapid changes and frequent critical situations. Humanity faces increasingly complex challenges, such as pandemics, wars, which can lead to social tensions. Having means for monitoring and measuring tension allows both society and the state to respond in a timely and adequate manner to these challenges. Tension leads to social conflicts, political crises, and economic downturns. Having a system for measuring tension helps us understand what factors contribute to tension and take steps to prevent or mitigate the effects. Therefore, it is urgent to develop tools to measure societal tensions, as this is an important step in the direction of understanding and managing social dynamics.
The purpose of the article. The purpose of this paper is to apply the methods of machine learning and neurolinguistic programming to the task of analyzing the opinions of Internet users to predict social tension in society.
Results. In this work, for the analysis of the problems of social dynamics, it is proposed to combine the approach using the vector representation of words and the clustering model in order to most accurately meet the needs of the developed program, which operates on open, unobserved text data in the Ukrainian language. The architecture and software of the social dynamics system based on machine learning methods were developed. It is divided into four modules: text data processing, Word2Vec model training, K-Means model training and user interface; models were trained with different manually adjusted hyperparameters. A graph of social tensions is presented, showing trends in the social dynamics of Ukrainians.
Keywords: tension analysis, social dynamics, machine learning, text data processing, word2vec, k-means.
Cite as: Tretynyk V., Nad Y. Using Machine Learning Methods to Develop a System of Social Dynamics. Cybernetics and Computer Technologies. 2025. 1. P. 81–88. (in Ukrainian) https://doi.org/10.34229/2707-451X.25.1.8
References
1. Harshali P.P., Atique M. Sentiment Analysis for Social Media: A Survey. 2nd International Conference on Information Science and Security (ICISS), 14-16 December 2015, Seoul, Korea (South). https://doi.org/10.1109/ICISSEC.2015.7371033
2. Kapil P., Ekbal A., Das D. Investigating Deep Learning Approaches for Hate Speech Detection in Social Media. International Conference on Computational Linguistics and Intelligent Text Processing, 2019. https://doi.org/10.48550/arXiv.2005.14690
3. Macis L., Tagliapietra M., Meo R., Pisano P. Breaking the trend: Anomaly detection models for early warning of socio-political unrest. Technological Forecasting and Social Change. Vol. 206. 2024. https://doi.org/10.1016/j.techfore.2024.123495
4. Redl C., Hlatshwayo S. Forecasting Social Unrest: A Machine Learning Approach. International Monetary Fund. 2021. Vol. 2021, Iss. 263. P. 29. https://doi.org/10.5089/9781557758873.001
5. Shchoholiev M., Tretynyk V. The System of Determining the Overall Estimate of the Level of Tension in Society. Digital Transformation, Cyber Security and Resilience of Modern Societies, Studies in Big Data. Vol. 84. P. 475–485. https://doi.org/10.1007/978-3-030-65722-2
6. Dataminr: About us. https://www.dataminr.com/about (accessed: 30.01.2025)
7. Crisis Monitor. https://www.cbs.nl/en-gb/about-us/innovation/project/social-tensions-indicator-gauging-society (accessed: 30.01.2025)
8. Brandwatch Overview. https://www.brandwatch.com/ (accessed: 30.01.2025)
ISSN 2707-451X (Online)
ISSN 2707-4501 (Print)
Previous | FULL TEXT (in Ukrainian) | Next