Research on the Construction of Real-Time Monitoring and Dynamic Evaluation Index System for Teaching Process Supported by Big Data Technology

Current educational evaluation systems often suffer from fragmented data collection, single-dimensional assessment, and delayed feedback, making it difficult to perceive learning processes in real time or provide timely intervention. This study develops a real-time monitoring and dynamic evaluation index system for teaching processes supported by big data technology. The Delphi method is used to select indicators, and the Analytic Hierarchy Process determines weights for a multidimensional model covering learning behavior, cognitive development, affective attitude, and social interaction, with 23 secondary indicators. A Hadoop distributed storage architecture, Spark real-time computing engine, and machine-learning models are used to process large-scale teaching data and identify learning states. Empirical research across three universities and 1,247 students shows that the system achieves 97.3% data collection accuracy under high-concurrency conditions, with video behavior data reaching 98.6% accuracy and text input reaching 95.1%. The system provides precise support for teaching intervention and is applicable to wireless sensing networks, smart classrooms, and electromagnetic-compatible educational monitoring environments.