Beyond traditional analysis: Using machine learning to investigate intellectual capital disclosures

Abstract

This research aims to conduct a thorough examination of the practices related to the disclosure of intellectual capital (IC). In the context of the dynamic knowledge-based economy, it is crucial for organizations to recognize the significance of IC. Intellectual capital comprises three key components: internal capital, external capital, and human capital. These elements play a pivotal role in generating value for organizations and positioning them competitively in the market. The current body of literature is constrained in its empirical investigation of IC disclosures and their congruence with organizational strategies. This research utilizes a combination of textual analysis and machine learning techniques, specifically K-means clustering, to examine the practices of IC disclosure. The integration of machine learning techniques facilitates the identification of patterns and interdependencies among diverse IC attributes. Notably, while the current literature has predominantly focused on IC disclosures within established frameworks, it often falls short of empirically exploring patterns and interconnections between different IC attributes. The results of the study indicate a notable emphasis on the disclosure of human capital and provide valuable insights into the different strategic priorities based on the clustering of IC attributes. The insights provided offer significant value to organizations as they facilitate the improvement of transparency and the effective communication of the strategic importance of IC. Furthermore, this study makes a valuable contribution to the existing theoretical framework on IC by identifying the interconnections that exist between various attributes of IC. The utilization of these findings by policymakers and standard-setting bodies can be instrumental in the development of more extensive guidelines for IC disclosures.