Abstract & Overview
The purpose of the research is to build a machine learning model for the decomposition of the textual formulation of competence. This paper discusses methods for representing competence as a set of keywords and terms. The main content of the research is the analysis of the application of two models: BERTopic and ARTM. The main stages of the research are described: data collection and preprocessing, model training, analysis and interpretation of the result. In conclusion, the advantages and disadvantages of using such models, as well as subsequent directions of research, are revealed.