Efficiency evaluation of scientific data sharing platforms in serving research considering heterogeneity based on DEA: a case study of national science data centers in China.

This article focuses on evaluating the efficiency of scientific data sharing platforms in supporting research, using China's national science data centers as a case study. It introduces a novel two-stage efficiency evaluation model based on Data Envelopment Analysis (DEA) that accounts for the platforms' roles in both data collection/sharing and application/transformation, while addressing disciplinary heterogeneity through an adjustment coefficient method suitable for small sample sizes. The study finds significant disparities in efficiency among the centers, with generally higher efficiency in data collection and sharing than in promoting data application, and recommends differentiated resource optimization, subject-specific data service strategies, and classification-based benchmarking to improve platform performance. The framework offers a platform-oriented perspective on research evaluation, highlighting the importance of technological and managerial improvements for enhancing scientific data sharing infrastructures.