A data-driven framework for assessing structural heterogeneity and nonlinear performance in university sports governance

University sports governance is a complex system constrained by multiple competing objectives, including resource allocation, participation equity, operational efficiency, and sustainable development. Traditional evaluation paradigms, which rely heavily on linear assumptions and weighted aggregation, are fundamentally inadequate for capturing the structural heterogeneity and nonlinear dynamics inherent in these organizations. To address this limitation, this study proposes a data-driven computational framework for heterogeneity assessment and performance mapping. Using Monte Carlo simulation, we generated 348 synthetic institutional samples and applied K-means and Gaussian Mixture Models (GMM) to assess whether the proposed pipeline can recover four pre-specified governance archetypes: resource-oriented, equity-prioritized, efficiency-driven, and balanced. Simulation results confirm the framework's technical identifiability and internal consistency under controlled conditions. Importantly, the configurations, performance patterns, and governance insights reported here reflect properties of the simulation design rather than empirically validated regularities in real university sports governance systems. Future research should apply this framework to actual institutional data to evaluate its external validity and practical relevance.