Analysis of Changes in Employee Sentiments Over Time Using Text Mining: A Case Study of Large-Scale Mergers and Acquisitions in the Japanese Petroleum Industry

This study examines how Mergers and Acquisitions (M&A) affect employee sentiment in Japan's petroleum industry by comparing differences between managers and general employees. We analyzed ۱۰,۳۷۸ open-ended survey responses collected over two years (۲۰۱۹: ۵,۱۱۷ responses, ۶۴.۹۶% response rate; ۲۰۲۰: ۵,۲۶۱ responses, ۴۷.۱۸% response rate) using natural language processing and sentiment analysis to track changes in employee sentiment after integration. Findings show that managers experienced increased anxiety and stress due to information overload, whereas general employees generally maintained positive attitudes and appreciated better communication opportunities with management. Both groups shared concerns about human resource evaluation systems and organizational structures. However, the specific nature and intensity of these concerns varied by role. These results indicate that uniform information-sharing and support are insufficient for successful Post-Merger Integration (PMI). Instead, organizations need targeted approaches that consider the different roles and information-processing patterns of each employee group. This study contributes a theoretical framework that combines information processing theory with M&A organizational dynamics to explain how hierarchical positions shape emotional responses during integration. Findings offer practical recommendations for human resource managers by emphasizing the importance of regular sentiment monitoring, role-specific communication strategies, and a clear demonstration of integration benefits to improve PMI outcomes.

کلیدواژه ها:

Employee sentiment ، Mergers and acquisitions (M&A) ، Post-merger integration (PMI) ، text mining ، Petroleum industry