Received 02.01.2025, Revised 11.03.2025, Accepted 24.04.2025

Method for constructing a cognitive map of processes in a dynamic system using the cooperation of large language models

Borys Varer, Vitalii Mokin

In the context of growing demands for rapid decision-making and in-depth analysis of complex dynamic systems – particularly when available data are limited and the involvement of experienced experts is either impractical or prohibitively expensive – the development of new methods for the construction of the model becomes especially relevant. The use of large language models (LLMs) as expert systems offers significant reductions in resource expenditure and accelerates the modelling of complex technical, environmental, and socio-economic systems. This study aimed to investigate and demonstrate the potential and capabilities of LLMs as expert systems in constructing cognitive maps. The article proposes and substantiates an architecture for the cooperation of LLM ensembles to formally generate vertices-variables and weight coefficients in cognitive maps, thereby enabling the automation of the modelling process without the involvement of human experts. A typical prompt for an LLM was decomposed into structural components: context description (D), model role instruction (R), instruction (I), conditions (C), and response format (F). A method for determining these components through expert-based analysis is proposed. A prompt system was developed to enable structured data processing and the identification of interrelationships among system elements. The practical effectiveness of the approach was demonstrated using a case study on forecasting water quality in the Sabarivske Reservoir near Vinnytsia. For most physicochemical indicators, the modelling showed low error rates (2.09-4.6%), even with a minimal amount of input data. The proposed method is promising for modelling and forecasting tasks in complex systems with limited data availability, particularly in environmental, socio-economic, and engineering contexts, where the speed of obtaining reliable results is critical for informed decision-making

LLM; generative artificial intelligence; intelligent technology; systems analysis; modelling; forecasting; dynamic system
69-78
Varer, B., & Mokin, V. (2025). Method for constructing a cognitive map of processes in a dynamic system using the cooperation of large language models. Information Technologies and Computer Engineering, 22(1), 69-78. https://doi.org/10.63341/vitce/1.2025.69

References

[1] Akiba, T., Shing, M., Tang, Y. & Ha, D. (2025). Evolutionary optimization of model merging recipes. Nature Machine Intelligence, 7(2), 195-204. doi: 10.1038/s42256-024-00975-8.

[2] Bender, E.M., & Koller, A. (2020). Climbing towards NLU: On meaning, form, and understanding in the age of data. In Proceedings of the 58th annual meeting of the association for computational linguistics (pp. 5185-5198). Stroudsburg: Association for Computational Linguistics. doi: 10.18653/v1/2020.acl-main.463.

[3] Bender, E.M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? In Proceedings of the 2021 ACM conference on fairness, accountability, and transparency (FAccT‘21) (pp. 610-623). Stroudsburg: Association for Computing Machinery. doi: 10.1145/3442188.3445922.

[4] Cao, H., Ma, R., Zhai, Y., & Shen, J. (2024). LLM-Collab: A framework for enhancing task planning via chain-of-thought and multi-agent collaboration. Applied Computing and Intelligence, 4(2), 328-348. doi: 10.3934/aci.2024019.

[5] Chen, S.-W., & Hsu, H.-J. (2023). MisCaltral: Reducing numeric hallucinations of mistral with precision numeric calculation. Research Square. doi: 10.21203/rs.3.rs-3789011/v1.

[6] Cherniuk, O. (2023). Modeling the impact of AI-based chatbots on the quality of higher education using system analysis methods. Information Technologies and Society, 3(9), 80-90. doi: 10.32689/maup.it.2023.3.11.

[7] Das, S., & Srihari, R. (2024). Compos mentis at SemEval2024 Task6: A multi-faceted role-based large language model ensemble to detect hallucination. In Proceedings of the 18th international workshop on semantic evaluation (SemEval-2024) (pp. 1449-1454). Mexico City: Association for Computational Linguistics. doi: 10.18653/v1/2024. semeval-1.208.

[8] Feleki, A., Apostolopoulos, I.D., Moustakidis, S., Papageorgiou, E.I., Papathanasiou, N., Apostolopoulos, D., & Papandrianos, N. (2023). Explainable deep fuzzy cognitive map diagnosis of coronary artery disease: Integrating myocardial perfusion imaging, clinical data, and natural language insights. Applied Sciences, 13(21), article number 11953. doi: 10.3390/app132111953.

[9] Godoy, W.F., Fabri, J.A., Palácios, R.H.C., Mendonça, M., Gonçalves, J.F.S., & Moraes, L.O.M. (2024). Using fuzzy cognitive maps and chatbots to evaluate student satisfaction in a university: A comparison between strong and weak AI. In Proceedings of the eighteenth international conference on mobile ubiquitous computing, systems, services and technologies (pp. 16-20). Wilmington: IARIA Press.

[10] Huang, L., et al. (2025). A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions. ACM Transactions on Information Systems, 43(2), article number 42. doi: 10.1145/3703155.

[11] Kong, A., Zhao, S., Chen, H., Li, Q., Qin, Y., Sun, R., Zhou, X., Wang, E., & Dong, X. (2024). Better zero-shot reasoning with role-play prompting. In Proceedings of the 2024 conference of the North American chapter of the association for computational linguistics: Human language technologies (Vol. 1, pp. 4099-4113). Mexico City: Association for Computational Linguistics. doi: 10.18653/v1/2024.naacl-long.228.

[12] Liu, T.J., Boulle, N., Sarfati, R., & Earls, C. (2024). LLMs learn governing principles of dynamical systems, revealing an in-context neural scaling law. In Proceedings of the 2024 conference on empirical methods in natural language processing (pp. 15097-15117). doi: 10.18653/v1/2024.emnlp-main.842.

[13] Lu, J., Pang, Z., Xiao, M., Zhu, Y., Xia, R., & Zhang, J. (2024). Merge, ensemble, and cooperate! A survey on collaborative strategies in the era of large language models. ArXiv. doi: 10.48550/arXiv.2407.06089.

[14] Mokin, V.B., Burdeina, O.V., & Varchuk, I.V. (2020). On the optimization of topologically observable cognitive maps while preserving their robustness. Visnyk of Vinnytsia Polytechnic Institute, (6), 84-92. doi: 10.31649/1997-9266-2020153-6-84-92.

[15] Mokin, V.B., Dratovanyi, M.V., Kozachko, O.M., & Zhukov, S.O. (2021). Method for synthesizing a robust multi-connected cognitive map of a complex system. Visnyk of Vinnytsia Polytechnic Institute, (6), 114-122. doi: 10.31649/1997-92662021-159-6-114-122.

[16] Roberts, F. (1976). Discrete mathematical models with applications to social, biological, and environmental problems. Englewood Cliffs: Prentice-Hall.

[17] Romanenko, V., & Miliavskyi, Y. (2022). Coordinating control of a cognitive map impulse process in stochastic environment. Problems of Control and Informatics, 67(4), 49-58. doi: 10.34229/2786-6505-2022-4-4.

[18] Saliieva, O., & Yaremchuk, Yu. (2020). Study of the reliability of the impact of threats on the level of security of the information protection system and the object of critical infrastructure based on the results of cognitive modeling . Bulletin of Cherkasy State Technological University, 25(3), 85-93. doi: 10.24025/2306-4412.3.2020.216251.

[19] Schoenegger, P., Tuminauskaite, I., Park, P.S., Bastos, R.V.S., & Tetlock P.E. (2024). Wisdom of the silicon crowd: LLM ensemble prediction capabilities rival human crowd accuracy. Science Advances, 10(45), article number eadp1528. doi: 10.1126/sciadv.adp1528.

[20] Schuerkamp, R., Ahlstrom, H., & Giabbanelli, P.J. (2025). Automatically resolving conflicts between expert systems: An experimental approach using large language models and fuzzy cognitive maps from participatory modeling studies. Knowledge-Based Systems, 313, article number 113151. doi: 10.1016/j.knosys.2025.113151.

[21] Shevchenko, S., Zhdanova, Y.A., Kryvytska, O., Shevchenko, H., & Spasiteleva, S. (2024). Fuzzy cognitive mapping as a scenario approach for information security risk analysis (short paper). In Proceedings of the 2024 cybersecurity providing in information and telecommunication systems II (pp. 356-362). Kyiv: Borys Grinchenko Kyiv Metropolitan University.

[22] Vinnytsia City Council. (2024). Ecology and natural resources. Retrieved from https://www.vmr.gov.ua/ecology.

[23] Wang, Z.M., et al. (2024). RoleLLM: Benchmarking, eliciting, and enhancing role-playing abilities of large language models. In Findings of the Association for computational linguistics (ACL 2024) (pp. 14743-14777). Bangkok: Association for Computational Linguistics. doi: 10.18653/v1/2024.findings-acl.878.