Topic Modeling of Employee Value Propositions: a Case of Russian IT Specialists via Crowdsourcing Platforms

This master's thesis studies how employee features are related to the perception of Employee Value Propositions (EVPs) in the Russian IT industry. The research is based on the data from the "Habr Career" platform and reviews gathered between 2018 and 2024 are analyzed using advanced topic modeling and sentiment analysis methods. The research aims to find out the general meanings and feelings in employee reviews, how these change among different job roles and experience levels, and then provide practical suggestions for organizations to improve their EVPs. The research's technique comprises data pre-processing, topic modeling using the BERTopic tool, and sentiment analysis with the DistilBERT model. The results of the research indicate that technology and market salaries, decision-making processes, and work-life balance are the key themes. They show how crucial it is to have tailored EVPs to improve employee satisfaction and retention. The thesis is a part of the theoretical foundation for EVPs, which combines Resource-Based View (RBV) and Human Capital (HC) theories, and also it shows how to use automated analysis in HR management.