I.D. Strebkov1

1 Federal Research Center "Computer Science and Control" of the RAS (Moscow, Russia)

1 istrebkov@frccsc.ru

Problem Statement. Knowledge graphs of semantic linked data are often characterized by incomplete links and the presence of isolated nodes, which reduces search efficiency and complicates their use for training language models. Analyzing the completeness and correctness of a knowledge graph is an open problem in assessing the knowledge quality of semantic libraries.

Goal. To use metric analysis methods for the knowledge graph of mathematical domains in the SciLibAIRU library to evaluate its structure, identify key and isolated nodes, and construct embeddings that improve navigation and search.

Results. Graph metrics were obtained, isolated concepts were identified, and embeddings were constructed to visualize its structure. Thresholds were established for identifying significant nodes, and a graph filtering method was proposed.

Practical significance. This approach enables improvement of the knowledge graph by replenishing links and optimizing navigation, which improves search accuracy and the quality of datasets for training domain-specific language models.

Strebkov I.D. Metric tools for analyzing the knowledge graph of subject areas in a semantic library. Highly Available Systems. 2026. V. 22. № 1. P. 95−98. DOI: https://doi.org/10.18127/j20729472-202601-19 (in Russian)

1. West R., Gabrilovich E., Murphy K. et al. Knowledge base completion via search-based question answering. 2014. WWW. 515–526.
2. Barbosa D., Wang H., Yu C. Shallow information extraction for the knowledge web. 2013 IEEE 29th International Conference on Data Engineering (ICDE). IEEE, 2013. P. 1264-1267. https://doi.org/110.1109/ICDE.2013.6544920
3. Ataeva O.M., Tuchkova N.P., Tejmurazov K.B., Abdy`shov A., Kobuk M.G. Biblioteka nauchny`x predmetny`x oblastej SciLibRu // E`lektronny`e biblioteki. 2025. T. 28. Vy`p. 6. S. 1324–45. https://doi.org/10.26907/1562-5419-2025-28-6-1324-1345
4. Ataeva O., Serebryakov V., Tuchkova N. Ontological approach to a knowledge graph construction in a semantic library. Lobachevskii J. of Mathematics. 2023. 44. P. 2229–2239. https://doi.org/10.1134/S1995080223060471
5. Park N., Kan A., Dong X. L., Zhao T. Christos Faloutso 2019. Estimating Node Importance in Knowledge Graphs Using Graph Neural Networks. In The 25th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD ’19), August 4–8, 2019, Anchorage, AK, USA. ACM, New York, NY, USA, 11 p. https://doi.org/10.1145/3292500.3330855
6. Gerasimenko N.A., Vatolin A.S., Yanina A.O., Voronczov K.V. SciRus: legkij i moshhny`j mul`tiyazy`chny`j e`nkoder dlya nauchny`x tekstov. Dokl. RAN. Matem., inform., procz. upr. 2024. 520:2. S. 216–227. Dokl. Math., 110:suppl. 1 (2024), S193–S202 https://doi.org/10.1134/S1064562424602178