Optimizing Student Development Pathways and Personal Competency Enhancement Strategies in Colleges and Universities Using Cluster Analysis

Abstract

The development of higher education is inseparable from the cultivation and development planning of students. The current quality of higher education development in universities and the rationality and effectiveness of student development planning directly affect the effective value of higher education in serving society. Through empirical analysis and the application of data-driven methods, this study optimizes students' future development pathways, particularly focusing on the development characteristics of undergraduate students from the 2020-2023 cohorts at a comprehensive higher education institution. The study employs an optimized extended centroid clustering algorithm combined with a Bayesian scoring function to classify the clustering results derived from multivariate data—including student growth pathways, participation in extracurricular activities, and mental health status—into five distinct categories with significant differences. Through research experiments, compared to traditional extended centroid clustering methods, the co-word clustering algorithm developed in this study can more clearly demonstrate the developmental characteristics of student groups. The co-word clustering method utilizes keyword data mining to extract sample characteristics of students' multidimensional development, thereby conducting multi-angle clustering to analyze the basic characteristics of student types and providing recommendations for fostering differentiated developmental capabilities among students.