In the midst of the global digital transformation, Central China Normal University (CCNU) is committed to becoming a world-class, China-characteristic benchmark university in “Artificial Intelligence + Education.” The university is fully advancing its digital education transformation strategy, leveraging digital technologies to build a top-tier institution. Guided by the principles of “data-driven development and integrated innovation,” CCNU leads pioneering efforts in education digital transformation through its “Digital CCNU” initiative. By continuously forging its unique path, CCNU is strengthening the cultivation of outstanding digital educators and shaping a distinctive “CCNU experience.”

As the 2025 World Digital Education Conference approaches—with CCNU hosting a parallel session on “Teacher Role Transformation and Capacity Enhancement in the Intelligent Era”—the university’s official WeChat account has launched a special column, “Digital Education at CCNU,” to highlight its explorations and successes in harnessing digital technologies for high-quality development.

For years, quality monitoring of digital course resources has largely depended on expert reviews and manual feedback, which often suffer from coarse evaluation granularity, limited coverage, and inconsistent standards. The sheer volume and diversity of course resources challenge traditional manual screening methods, necessitating an intelligent overhaul of existing management models. To address this, CCNU, supported by the National Engineering Research Center for Digital Learning, has innovatively integrated large language models and multimodal analysis techniques to develop a data-evidence-based approach to evaluating course resource quality. The system collects and analyzes multimodal data—including audio, video, and text—to establish a multi-level evaluation framework that marks a decisive shift from “experience-based judgment” to “data-driven decision-making,” providing intelligent support for resource evaluation and management in universities.

Large Model-Driven Multimodal Analysis: Enabling AI to Truly “Understand” a Lesson

Building on a thorough analysis of key factors in course resource quality, the research team designed an evaluation framework encompassing teaching content, instructional management, learning activities, classroom interaction, and technical support. Leveraging multimodal data analysis and large model technology, they developed an intelligent digital course resource evaluation system that delivers comprehensive insight and smart assessment of educational materials.

The system identifies critical quality features in real-time, precisely distinguishing high-quality resources from those with potential risks. It then generates targeted recommendations for improvement, driving ongoing optimization and enhancement of resources. By seamlessly combining dynamic monitoring, accurate attribution, and impact evaluation, the system creates a closed-loop mechanism for continuous quality improvement, injecting fresh momentum into the pursuit of high-quality education and teaching.

Widespread Implementation Empowering Modern Education Governance

To date, the system has been deployed in over 100 schools nationwide, completing more than 40,000 intelligent evaluations of course resources. It has significantly contributed to digital education reforms in regions such as Hubei and Ningxia. Thanks to the system’s real-time monitoring and precise analysis capabilities, education administrators can dynamically track resource status, swiftly identify outstanding and potentially problematic courses, and promote focused teaching research, teacher capacity building, and ongoing resource refinement. This establishes a new model of resource governance based on “real-time monitoring—precise diagnosis—dynamic improvement,” supporting universities in cultivating a high-quality educational ecosystem.