Research on Multi Source Data Fusion Evaluation System Based on Modular Knowledge Graph: Taking the Course of Database Principles and Applications in Colleges and Universities as an Example

Abstract

In view of the pain points of single evaluation, lack of process supervision, and weak practical evaluation of the course "Database Principles and Applications" in Colleges and universities, this study constructed a smart evaluation system based on modular knowledge map and integrating multi-source data (online behavior, experimental reports, project results, and enterprise evaluation), so as to achieve accurate portrait in the whole learning cycle. After a semester of controlled experiment (242 students), the experimental group was significantly better than the control group in the level of knowledge structure, engineering practice ability and autonomous learning engagement (p<0.05), which effectively solved the dilemma of "emphasizing results over process, emphasizing theory over practice". The system provides a replicable practice paradigm for the digital and accurate evaluation of computer courses in local private colleges and universities.