UDC 51-77
DOI: 10.36871/2618-9976.2023.06.002
Authors
Irina E. Kalina,
Master's Student, Research Engineer, Ural Federal University, Ekaterinburg, Russia
Yulia D. Sokolova,
Master's Student, Research Laboratory Assistant, Ural Federal University, Ekaterinburg, Russia
Alexey V. Shevchuk,
Master's Student, Assistant, Ural Federal University, Ekaterinburg, Russia
Oleg S. Mariev,
Institute of Economics, The Ural Branch of Russian Academy of Sciences, Ekaterinburg, Russia
Abstract
This paper assesses the degree of news coverage by topic by applying structural topic modelling. The preliminary stage involves the computeraided gathering and systematisation of information from the Lenta.ru online news portal, converting words into a single form, and cleaning the text data from stopwords. The method of structural thematic modelling (STM) was used in this study, which made it possible to identify the 25 most popular topics for publication on the news portal and to determine the frequency of occurrence of the words in the subject and their uniqueness. Correlation and regression analyses indicated the correlation of the topics under study and the change in the degree of coverage of the topics over time.
Keywords
Topic modelling, News, News portal, Correlation analysis, Regression analysis