Psychological Profiling of Users Based on Social Media Data Analysis Using the BERT Model

The article deals with the problem of automated construction of psychological portrait of social network users using machine learning and neural networks methods. In the conditions of the modern labour market, effective human resource management requires fast and accurate profiling of candidates. Traditional methods of personality analysis require significant time and human resources, which makes the process time-consuming and not always prompt. The importance of psychoanalysis for human resource management is substantiated and the methodology of using MBTI typology as a tool for classifying personality characteristics is presented. The authors analyze modern research in the field of automated psychological profiling and consider the use of machine learning to analyze the texts of social network users. The paper presents the methodology of data collection and processing using APIs of social platforms, as well as the application of algorithms based on the MBTI model. Special attention is paid to the BERT method, which allows taking into account the context of words and providing high classification accuracy. The results of the study demonstrate high accuracy of personality type classification, reaching 97.72%, which confirms the effectiveness of the automated approach to profiling. The use of the developed model significantly reduces the analysis time and increases its predictive power. In the future, this area may become an important tool for HR-specialists, allowing them to optimize the recruitment process and analyze the personality characteristics of candidates.