Auxiliary Teaching Design for Knowledge Points Labeling by Education Data Analysis

With the in-depth integration of information technology and human life, the Internet has gained rapid popularity, and global data shows the characteristics of explosive growth and massive aggregation. The outbreak of COVID-19 makes online learning popular and universal. Different from the traditional way of education, online learning allows learners to have access to complexity and diversity as well as varying levels of knowledge mastery. For the current mainstream personalized education, making good use of education data for learning analysis is an important direction to solve the problem. This paper finds that learning platform has just begun to rise on a large scale, but there are still some problems, such as the confusion of learning resources, the unreasonable setting of the question bank and the unclear classification of the difficulty of knowledge points. Most of the platforms do not divide the knowledge points and difficulty of the questions, therefore, students cannot find the key points, and some of them fill in the difficulty directly according to subjective judgment, which is not scientific nor objective. This paper analyzes students' historical answer data, classifies and divides the difficulty of knowledge points by clustering algorithm, then verifying it on the Istudy education platform which is developed by our laboratory. The experimental results show that this labeling system is more objective and effective, contributing to improving the efficiency of online learning. For students, they can adjust their own learning plans while understanding the difficulty of knowledge points. Teachers can also modify their teaching focus and assist the teaching design according to the difficulty classification of knowledge points.