[2I6-GS-10] AI application: Finance

[2I6-GS-10-03] Text mining analysis of human capital information disclosure in a corporate governance report

Sei Kamitani¹, Miyuki Matsuoka¹, 〇Takumi Nakai¹ (1. Osaka City University)

Keywords:human capital, accounting, text mining

In this study's object, we visually analyzed the qualitative information on human capital from the "Corporate Governance Reports" submitted by companies listed on the Tokyo Stock Exchange through text mining. The mandatory disclosure of human capital information for approximately 4,000 major firms from the financial statements of March 2023 indicates a growing interest in corporate human capital information from stakeholders. Given the recent start of this obligation, the study is considered highly innovative. We conducted morphological analysis on the report, followed by extracting words with high relevance to human capital and analyzing them using hierarchical cluster analysis and co-occurrence network graphs. The hierarchical cluster analysis revealed strong associations with words like "women," "management," and "recruitment," while the co-occurrence network graph showed the coexistence of words such as "women," "empowerment," "position," and "power." This suggests an emphasis on nurturing and promoting female talent in the human capital investment strategy outlined in the reports. Additionally, it is believed that the stakeholders now have clearer points of interest in the report.