Integrating Text Mining With DEA-Malmquist Index for Evaluating First-Class Online Course Development Efficiency

The dynamic evaluation of development performance in nationally recognized first-class courses holds broad implications for improving the quality of online education. This study examines 512 nationally recognized first-class undergraduate online courses on iCourse platform, framing them as an input–output system and drawing on learner-generated online review data. By integrating text mining techniques with data envelopment analysis and the Malmquist index, the study dynamically assesses their development efficiency from 2019 to 2024. The findings indicate that (1) first-class course development maintained relatively high overall efficiency, although sustainability was constrained by limited post-approval investment; (2) course performance displayed a decline–rebound trajectory, primarily driven by teaching innovation; and (3) significant disciplinary differences emerged, with natural science courses exhibiting greater volatility but stronger resilience, whereas humanities and social science courses showed milder fluctuations and experienced late-stage innovation fatigue and resource aging.