A Multidimensional Text Mining Approach to Analyze Research Trends on Scalp and Hair in Korea

This research investigates the evolving trends in scalp and hair research in Korea through a text mining-based analytical framework. A total of 201 domestic academic articles published between 1994 and 2024 were analyzed using computational tools including Multidimensional Scaling (MDS), Self-Organizing Maps (SOM), co-occurrence network analysis, and time-series relative frequency analysis. The data were divided into two temporal phases: P1 (1994–2014) and P2 (2015–2024), to capture conceptual and structural changes over time. The MDS results indicated a transition from a single-cluster structure in P1 to a diversified, multi-nucleus keyword distribution in P2, reflecting an expansion of research focus into industrial, medical, and psychological domains. SOM analysis revealed that while P1 featured clearly segmented clusters centered on experimental and educational themes, P2 showed an integrated convergence of topics, suggesting blurred disciplinary boundaries and enhanced practical relevance. Co-occurrence network analysis demonstrated increased network density and cohesion in P2, implying more interconnected and complex research relationships. Furthermore, time-series analysis revealed a shift from closed, experimental approaches in early research to interdisciplinary, application-oriented studies in recent years. These findings highlight the academic evolution of scalp and hair research in Korea, indicating a transition from technique-driven investigation to a more systemic and consumer-responsive research paradigm. This study offers foundational insights for expanding interdisciplinary collaborations and guiding future developments in the beauty and wellness sectors.