AI for Workplace Insights: A Scalable NLP Approach to Identifying Employee Values on LinkedIn
Guy Doytch, Lev Muchnik · Academy of Management Proceedings · 2025
Understanding personal values is critical for interpreting employee behavior, decision-making, and organizational performance. Advances in Natural Language Processing (NLP) and the growing availability of online employee data offer unprecedented opportunities for studying values on a global scale and gaining new insights into work-related behaviors. This study introduces and validates an AI-based methodology for identifying personal values from LinkedIn self-descriptions, utilizing a dataset of over 100M employees. Leveraging Generative AI and BERT-based models, our approach successfully replicates theoretical constructs from well-established values theories, predicts outcomes such as tenure and national political orientations, and outperforms existing methods. Furthermore, our work provides organizational researchers with a novel paradigm for combining Generative AI and BERT-based models to create scalable frameworks, complemented by a robust validation methodology for advanced NLP techniques. To our knowledge, this is the first study to apply cutting-edge AI methods to analyze employee self-descriptions on a global scale. The AI model and accompanying Python and R code will be made freely available to researchers.