RESEARCH OF METHODS FOR CONSTRUCTING CAUSAL GRAPH MODELS FOR COMPLEX SOCIO-HUMANITARIAN SYSTEMS
Abstract
Complex socio-humanitarian systems are a type of systems that are studied in sociology, anthropology,
economics, political science, and other humanities. These systems are characterized by the complexity of
interactions between their constituent elements, which can be both human (individuals, groups) and cultural,
social, economic and political aspects. For example, society as a socio-humanitarian system consists of various
elements, such as people, culture, institutions, values, and so on. They interact with each other, forming a
complex network of connections and influences that determines the behavior and development of society.
To better understand such systems, various approaches are used, including systems analysis, social network
theory, complexity theory, and other methods. These approaches help to identify the main patterns in the
functioning of complex socio-humanitarian systems and predict their development in the future. This article
discusses approaches to identifying cause-and-effect relationships and highlights the basic requirements for
building these connections in the context of complex socio-humanitarian systems that deal mainly with semistructured
information, often in the form of natural language and texts. The strengths and weaknesses of the
identified approaches were identified, and examples of the use of modern methods of constructing graphs on
various tasks were considered: identifying risks in business, analyzing social phenomena, identifying the
presence of causality in texts. The study showed that the most productive methods are machine learning, for
example, language models for extracting knowledge from text in combination with neural network technologies
and graph representations of knowledge. They require solid knowledge of mathematics, statistics and
programming, at least in Python, with the most impressive tool support for solving machine learning problems.
Also, identifying causality is based not only on correlation but also on other methods such as the
Granger test used for time series analysis.
References
Social Science Research, 2022, Vol. 108, 102799. ISSN 0049-089X.
2. Lahti H., Kulmala M., Hietajärvi L., Lyyra N., Kleszczewska D., Boniel-Nissim M., Furstova J., van
den Eijnden R., Sudeck G., Paakkari L. What Counteracts Problematic Social Media Use in Adolescence?
A Cross-National Observational Study, Journal of Adolescent Health, 2024, Vol. 74, Issue 1,
pp. 98-112. ISSN 1054-139X.
3. Asteriou Dimitros, Hall Stephen G. ARIMA Models and the Box–Jenkins Methodology, Applied
Econometrics. 2nd ed. Palgrave MacMillan, 2011, pp. 265-286. ISBN 978-0-230-27182-1.
4. Terence C. Mills. Time Series Techniques for Economists. Cambridge University Press, 1990. ISBN
978-0-521-34339-8.
5. Percival Donald B., Walden Andrew T. Spectral Analysis for Physical Applications. Cambridge University
Press. 1993. ISBN 978-0-521-35532-2.
6. Shumway R.H. and Stoffer D.S. Time Series Analysis and Its Applications: With R Examples. Springer,
2017. DOI: 10.1007/978-3-319-52452-8.
7. Gers F., Schraudolph N., Schmidhuber J. Learning precise timing with LSTM recurrent networks
(engl.), Journal of Machine Learning Research: journal, 2002, Vol. 3, pp. 115-143.
8. Greff K., Srivastava . K., Koutník Jan, Steunebrink Bas . , Schmidhuber Jürgen. LSTM: A Search
Space Odyssey, Neural and Evolutionary Computing, 2015
9. Gers F.A., Schmidhuber J. LSTM Recurrent Networks Learn Simple Context Free and Context Sensitive
Languages (engl.), IEEE Transactions on Neural Networks (engl.) russ.: journal, 2001, Vol. 12,
No. 6, pp. 1333-1340. DOI: 10.1109/72.963769.
10. Kaminski M., Ding M., Truccolo W.A., Bressler S.L Evaluating causal relations in neural systems:
Granger causality, directed transfer function and statistical assessment of significance, Biol Cybern,
2001, 85, pp. 145-57.
11. Ancona N., Marinazzo D., Stramaglia S. Radial basis function approaches to nonlinear Granger causality
of time series, Physical Review E, 2004, Vol. 70, 056221.
12. Arendt F., Mestas M. Suicide among soldiers and social contagion effects: An interrupted time-series
analysis, Social Science & Medicine, 2023, Vol. 320, 115747. ISSN 0277-9536.
13. Ding L., Chen J., Du P., Xiang Ya. Event causality identification via graph contrast-based knowledge
augmented networks, Information Sciences, 2024, Vol. 656, 119905.
14. ConceptNet. Режим доступа: Available at: https://conceptnet.io/ (accessed 28 March 2024).
15. Bond F., Paik K. A survey of wordnets and their licenses, In Proceedings of the 6th Global WordNet
Conference (GWC 2012). Matsue, 2012, pp. 64-71.
16. Bond F., Vossen P., McCrae John. CILI: the Collaborative Interlingual Index./ F. Bond, P. Vossen,
John McCrae, Christiane Fellbaum, In Proceedings of the 8th Global WordNet Conference
(GWC2016). Bucharest, 2016, pp 50-57.
17. Pre-training of deep bidirectional transformers for language understanding, Proceedings of the 2019
Conference of the North American Chapter of the Association for Computational Linguistics: Human
Language Technologies, (Vol. 1: Long and Short Papers), 2019, pp. 4171-4186.
18. Cheng F., Miyao Y. Classifying temporal relations by bidirectional lstm over dependency paths, Proceedings
of the 55th Annual Meeting of the Association for Computational Linguistics, 2017, Vol. 2:
Short Papers, pp. 1-6.
19. Hajime Sasaki, Motomasa Fujii, Hiroki Sakaji, Shigeru Masuyama. Enhancing Risk Identification
with GNN: Edge Classification in Risk Causality from Securities Reports, International Journal of Information
Management Data Insights, 2024, Vol. 4. Issue 1, 100217. ISSN 2667-0968.
20. Wu Lingfei, Cui Peng, Pei Jian; Zhao Liang. Graph Neural Networks: Foundations, Frontiers, and
Applications. Springer Singapore, 2022, 725 p.
