Hierarchical Clustering Based Teaching Reform Courses Examination Data Analysis Approach Applied in China Open University System

Abstract

China Open University system, as an exclusive university organization that is specialized in distance open education in China, adopts the network teaching way, and whose teaching network covered the whole country, so the system's teaching quality is increasingly attracting attention, and the teaching reform measures used to improve the teaching quality also have been taken. Apparently, utilizing data mining techniques to analyze the teaching reform course examination data is an effective method to check the effects of teaching reform measures. Clustering analysis, as an unsupervised learning, could find the rule hidden in the data completely according to the data itself. Hierarchical clustering, has the advantages of classification accurately, outliers detection easily, and doesn't need to preset the cluster number. So this paper proposes an examination data analysis approach based on hierarchical clustering algorithm to check the effects of teaching reform measures happened in China Open University system. This paper describes an implementation scheme based on hierarchical clustering, designed for teaching reform course examination data analysis, including algorithm design and application design. The effectiveness of proposed approach is verified by processing the practical examination data in China Open University system's teaching reform courses. The experimental results reveal the changing regulation of the examination data caused by teaching reform measures, and could be the objective basis for open education teaching reform.