N.V. Blokhin1, S.V. Makrushin2

1, 2 Financial University under the Government of the Russian Federation (Moscow, Russia)

1, 2 Center for Intelligent Analytical and Robotic Systems (Balashikha, Russia)

1 nvblokhin@fa.ru, 2 svmakrushin@fa.ru

The sectoral classification of legal entities is usually based on the All-Russian Classifier of Types of Economic Activity (ACTEA). ACTEA allows us to systematize and group companies according to their main field of activity. However, this classifier is poorly suited for the purposes of assessing the economic interconnection of companies, since it reflects the technological similarity of fields of activity and does not take into account the relationship between industries along the chain of product conversion. The company's industry affiliation is determined not only by its declared ACTEA code, but also by its actual economic and legal relationships with other companies. Using modern machine learning methods, it is possible to build vector representations (embeddings) of companies, taking into account, among other things, the relationships between companies.

Blokhin N.V., Makrushin S.V. Construction of company embeddings based on a vector representation of economic sectors and connections between companies. Neurocomputers. 2025. V. 27. № 2. P. 5–12. DOI: https://doi.org/10.18127/j19998554-202502-01 (in Russian)

1. Mikolov T., Chen K., Corrado G., Dean J. Efficient estimation of word representations in vector space. Proceedings of the International Conference on Learning Representations. 2013.
2. Song C., Raghunathan A. Information leakage in embedding models. Proceedings of the 2020 ACM SIGSAC Conference on Computer and Communications Security. 2020.
3. Grbovic M., Haibin C. Real-time personalization using embeddings for search ranking at Airbnb. Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2018.
4. Beam A., Kompa B., Schmaltz A., et al. Clinical concept embeddings learned from massive sources of multimodal medical data. Pacific Symposium on Biocomputing. 2018.
5. Madjiheurem S., Qu L., Walder C. Chord2Vec: Learning musical chord embeddings. 30th Conference on Neural Information Processing Systems. 2016.
6. Blokhin N.V., Makrushin S.V. Postroenie vektornogo predstavleniya otraslej ekonomiki s pomoshch'yu grafovykh nejronnykh setej. Informatsionno-izmeritel'nye i upravlyayushchie sistemy. 2023. № 5. S. 7–15. (in Russian)
7. Prikaz Rosstata ot 31.12.2014 № 742 (red. ot 04.02.2016) «O metodicheskikh ukazaniyakh po opredeleniyu osnovnogo vida ekonomi­cheskoj deyatel'nosti khozyajstvuyushchikh sub''ektov na osnove Obshcherossijskogo klassifikatora vidov ekonomicheskoj deyatel'nosti (OKVED2) dlya formirovaniya svodnoj ofitsial'noj statisticheskoj informatsii». (in Russian)
8. Zhou J., Cui G. Graph neural networks: A review of methods and applications. arXiv preprint arXiv:1812.08434. 2018.
9. Zhang S., Tong H., Xu J., Maciejewski R. Graph convolutional networks: a comprehensive review. Computational Social Net-works. 2019. V. 6. № 11.
10. Wang X., Bo D., Shi C., et al. A survey on heterogeneous graph embedding: Methods, techniques, applications and sources. 2020.
11. Klassifikatsiya [Elektronnyj resurs]. URL: http://www.machinelearning.ru/wiki/index.php?title=Klassifikatsiya (data obrashcheniya: 07.12.2023). (in Russian)
12. Potdar K., Pardawala T.S., Pai C.D. A comparative study of categorical variable encoding techniques for neural network classifiers. International Journal of Computer Applications. 2017. V. 175. № 4. P. 7–9.
13. Hamilton W.L., Zhitao Y., Leskovec J. Inductive representation learning on large graphs. Neural Information Processing Systems. 2017.
14. Kingma D., Ba J. Adam: A method for stochastic optimization. International Conference on Learning Representations. 2014.