Method of building embeddings of signs in deep learning problems based on ontologies

: pp. 189 - 197
Lviv Polytechnic National University, Lviv, Ukraine
Lviv Polytechnic National University, Ukraine

This paper investigates the problem of embedding features used in datasets for training neural networks. The use of embeddings increases the performance of neural networks, and therefore is an important part of data preparation for deep learning methods. Such a process is based on semantic metrics. It is proposed to use ontologies of the subject areas to which the corresponding feature belongs for embedding. This work developed such a method and investigated its use for the task of categorizing text documents. The research results showed the advantage of the developed method.

  1. Lytvyn V. V. (2011). Knowledge bases of intelligent decision support systems: monograph. Lviv: Publishing House of Lviv Polytechnic, 240 p.
  2. Vdovichenko A. V. (2002). Intelligent search systems. Classification and comparison. Artificial intelligence, IPSI “Science and education”, No. 3, 61–70.
  3. Strube M., Ponzetto S. (2022). WikiRelate! Computing semantic relatedness using Wikipedia. In Proceedings of the 21st National Conference on Artificial Intelligence. (AAAI 06). Boston, Mass., July 16–20, 2022. Access mode:
  4. Jarmasz M., Szpakowicz S. (2020). Roget’s Thesaurus and semantic similarity. In Proceedings of Conference on Recent Advances in Natural Language Processing (RANLP 2003). Borovets, Bulgaria, September, 212–219.
  5. Fellbaum C. (1998). WordNet: an electronic lexical database. MIT Press, Cambridge, Massachusetts, 423 p.
  6. Wu Z., Palmer M. (1994). Verb semantics and lexical selection. In Proc. of ACL-94, 133–138.
  7. Resnik P. (1995). Disambiguating noun groupings with respect to WordNet senses. In Proceedings of the 3rd Workshop on Very Large Corpora. MIT, June. Access mode:
  8. Resnik P. (2019). Semantic similarity in a taxonomy: an information-based measure and its application to problems of ambiguity in natural language. Journal of Artificial Intelligence Research (JAIR), Vol. 11, 95–130.
  9. Lin D. (2018). An information-theoretic definition of similarity. In Proceedings of International Conference on Machine Learning, Madison, Wisconsin, July. Access mode: http://www.cs
  10. WordNet: a lexical database for the English language. Cognitive Science Laboratory Princeton University, 2006. Access mode:
  11. Gruninger M., Fox M. (1995). Methodology for the Design and Evaluation of Ontologies. Proceedings of IJCAI-95 Workshop on Basic Ontological Issues in Knowledge Sharing, 231–238.
  12. WordNet: a lexical database for the English language. Cognitive Science Laboratory Princeton University, 2006. Access mode:
  13. Dubinsky A. G. (2001). Development of models and improvement of the structure of information search systems in the global computer network: abstract. dis... cand. technical sciences: 05.13.06 / NAS of Ukraine; National Library of Ukraine named after V. I. Vernadskyi. K., 17 p.
  14. Bulskov H., Knappe R., Andreasen R. (2004). On Querying Ontologies and Databases. FQAS, 191–202.
  15. Kravets P. O., Lytvyn V. V., Vysotska V. A. (2022). Simulation of the game task of assigning personnel for the execution of IT projects based on ontologies. Radio electronics, informatics, management, No. 1, 130–145.
  16. Bublyk M., Kowalska-Styczeń A., Lytvyn V., Vysotska V. (2021). The Ukrainian economy transformation into the circular based on fuzzy-logic cluster analysis. Energies, 14(18), 5951. Access mode:
  17. Kravets P., Lytvyn V., Vysotska V. (2020). Game Model of Ontological Project Support. Radio Electronics, Computer Science, Control, Vol. 1(1), 172–183. Access mode:
  18. Karpov I. A., Burov E. V. (2020). The use of ontological networks in decision support systems under conditions of ambiguity. Bulletin of the Lviv Polytechnic National University. Series: Information systems and networks, is. 7, 8–15. Access mode: ontologichnyh-merezh-u-systemah-pidtrymky-pryynyattya.