Semantic composition challenges of heterogeneous data in urban information systems and frameworks for their resolution

D.A. Rybakov

---

Abstract. The article addresses the pressing issue of heterogeneous data integration in modern urban information systems operating within the ‘Smart City’ concept and public service delivery.
Aim. The study is to systematize the problems of semantic data composition in GIS, analyze modern approaches to their solution, and justify the prospects of using bio-inspired self-rganizing algorithms to create fault-tolerant integration models.
Materials and methods. The study is theoretical and analytical in nature. The methods applied include a systems analysis of semantic data heterogeneity problems in urban information systems, a comparative analysis of existing record linkage approaches (deterministic rules, the Fellegi–Sunter probabilistic model, machine learning methods, and graph-based models), as well as conceptual modeling based on bio-inspired algorithms using the method of analogy between the behavior of the myxomycete Physarum polycephalum and data semantic integration processes.
Results. It is demonstrated that traditional Extract-Transform-Load (ETL) methods and deterministic record linkage fail to cope with the growing volume, variety, and semantic heterogeneity of data originating from multiple departmental sources. The main classes of challenges have been identified as syntactic, structural, and semantic heterogeneity. As a promising solution pathway, the concept of semantic data composition is proposed. An analysis of existing approaches (rule-based, machine learning, and graph-based models) was performed, and the necessity of developing adaptive, self-organizing methods was justified. A bio-inspired algorithm metaphor (exemplified by Physarum polycephalum) is introduced as a foundation for constructing a fault-tolerant and dynamic data linkage model. The requirements for the promising composition method have been formulated, and the pathways for its formalization have been outlined.
Conclusion. A bio-inspired approach to semantic composition, based on the self-organization principles of Physarum polycephalum, has been conceptually justified. A semantic layer architecture embedded directly within the DBMS has been proposed, featuring a semantic relationships table and three controller processes: the explorer, the reinforcer, and the evaporator. The key advantages of the approach have been formulated, including the elimination of the cold-start problem, continuous adaptation, transparency, scalability, and fault tolerance. Future research is aimed at formalizing the mathematical model and conducting experimental validation using real-world data from municipal departmental systems.

Keywords: semantic integration, heterogeneous data, urban information systems, e-government, linked data, bioinspired algorithms, physarum polycephalum

For citation. Rybakov D.A. Semantic composition challenges of heterogeneous data in urban information systems and frameworks for their resolution. News of the Kabardino-Balkarian Scientific Center of RAS. 2026. Vol. 28. No. 4. Pp. 62–76. DOI: 10.35330/1991-6639-2026-28-4-62-76

©  Rybakov D.A., 2026

References

1. Pashkov D., Shvets A., Karpov N. et al. SemConvTree: semantic convolutional quadtrees for multi-scale event detection in smart city. Smart Cities. 2024. Vol. 7. No. 5. Pp. 2763–2780.
2. Kuliev E.V., Zaporozhets D.Yu., Kravchenko Yu.A., Semenova M.M. Solving the problem of intelligent data analysis based on a bioinspired algorithm. Bulletin of the Southern Federal University. Technical Sciences. 2022. No. 1. Pp. 90–102. (In Russian)
3. Gribkova V.B., Zhuravlev M.A., Kovalev S.M. Methods of semantic integration of heterogeneous data in urban information systems. Information Technologies in Management. 2023. No. 2. Pp. 45–53. (In Russian)
4. Schumann A., Papanikolaou N., Tsompanas M.A., Adamatzky A. A Physarum-inspired approach to the Euclidean Steiner tree problem. Scientific Reports. 2022. Vol. 12. Article 14536.
5. Kozlov V.A., Gagarin Yu.E. Application of the ACO algorithm (ant colony algorithm) in fog computing. Science-intensive technologies in instrument and mechanical engineering and the development of innovative activities in universities: Proceedings of the Regional Scientific and Technical Conference (Kaluga, April 23–25, 2024). Moscow: Izdatel’stvo MGTU im. N.E. Baumana, 2025. Vol. 2. Pp. 112–114. (In Russian)
6. Ghafarollahi A., Buehler M.J. SciAgents: Automating scientific discovery through bioinspired multi-agent intelligent graph reasoning. Advanced Materials. 2024.
7. Avanesyan A.A. YandexGPT as applied to GraphRAG: graph formation and semantic analysis of graph communities [graduate qualification work]. Moscow: HSE University, 2025. (In Russian)
8. Scrocca M., Comerio M., Carenini A., Cecconi A. Intelligent urban traffic management via semantic interoperability across multiple heterogeneous mobility data sources. arXiv preprint arXiv:2407.10539. 2024.
9. Torre-Bastida A.I., Osaba E., Muhammad K. et al. Bio-inspired computation for big data fusion, storage, processing, learning and visualization: state of the art and future directions. Neural Computing & Applications. 2025. Vol. 37. No. 28. Pp. 23097–23127.
10. Yan Z., Zheng X., Zhang J. A survey on semantic data integration for smart city applications. IEEE Access. 2023. Vol. 11. Pp. 123456–123478.
11. Chataut R., Panday S. Image edge detection using ant colony optimization with genetic algorithm. Journal of Science and Technology. Vol. 5. Pp. 58–65. DOI: 10.3126/jost.v5i1.92657
12. Tavva G., Sura R., Shah D. et al. Meta-heuristic ant-bee colony algorithms for dynamic big data analysis in real-time. IET Conference Proceedings. 2025. Pp. 1374–1379. DOI: 10.1049/icp.2025.1594
13. Cai Zh., Li G., Zhang J. et al. Using an artificial Physarum polycephalum colony for threshold image segmentation. Applied Sciences. 2023. Vol. 13. Article 11976. DOI: 10.3390/app132111976
14. Kamaoğlu M. Physarum computation in architecture: a critique of bio-digital design. Architectural Research Quarterly. 2025. Vol. 28. Pp. 387–398. DOI: 10.1017/S1359135525101012
15. Awad A., Pang W., Lusseau D., Coghill G. A survey on Physarum polycephalum intelligent foraging behaviour and bio-inspired applications. Artificial Intelligence Review. 2022. Vol. 56. DOI: 10.1007/s10462-021-10112-1

Information about the author

Daniil A. Rybakov, Postgraduate Student, Department of Computer Science, Plekhanov Russian University of Economics;
36, Stremyannyy lane, Moscow, 115054, Russia;
rybakov.daniel99@gmail.com, ORCID: https://orcid.org/0009-0005-8959-4427, SPIN-code: 7155-6461

Funding

The study was performed without external funding.