Research on Course Evaluation Index Selection Based on Decision Tree Algorithm

Abstract

High-quality course teaching is the goal pursued by modern universities. The traditional course evaluation system has the characteristics of multiple indexes and fuzzy boundaries between indexes, which has defects such as redundancy and even invalid indexes. In order to improve the evaluation efficiency and enhance the user experience of the evaluation subject, this paper proposed a course evaluation index selection method based on decision tree. The course evaluation data of a university was selected as the research data, including10 indexes and 632 courses with a total of 138,635 records. After the preprocessing operations of summarizing and averaging on the research data, the discrete data was obtained by K-means clustering algorithm, and the corresponding classification label was obtained for each course. Then, a classification model was constructed based on decision tree algorithm C4.5. Two of the 10 indexes are filtered by the decision tree. The experimental results showed that the accuracy of our model reached 90.5%. Therefore, the method proposed could effectively improve the reliability of the course evaluation system.