The Research Of Groupingalgorithm Of Independent Study Resource Of Chinese Based On CURE Clustering

With the rapid development of Chinese economy, science and technology, Mandarin fever is experiencing continuous expansion around the world and the number of Chinese learners is increasing sharply. Since online teaching has the characteristics of fast, efficient and free, and it has already gotten rid of the limits of time and geographical factors of traditional teaching, it could make full use of multimedia features such as images, animation, video and so on to make the original abstract and complex learning contents become more abundant, more directly, and as a result, it helps improve the learning efficiency of Chinese learners. However, online teaching has its limitation. For example, the online learning contents are just accumulated and listed simply but not organized reasonably, which seriously affect the efficiency of network teaching as the learners with different learning backgrounds are difficult to distinguish which part fit for themselves.Resources play a very important role in the process of network autonomous learning; furthermore, the interest of Chinese learners is directly affected by the difficulty level of resources. Thus, the main issue of network autonomous learning is to find the suitable learning resources for learners.This paper makes clear the characteristics of network autonomous learning as well as points out the environmental factors affecting the efficiency of Chinese learners through analyzing the characteristics of network autonomous learning and Chinese learners, and makes clear the learners'requirements of resources by regarding learning resources of Chinese language as a clustering data set and basing on the properties, ages, genders, personalities, Chinese language levels, mother tongues and cultural backgrounds of students. Through learner's activity data, learning resources of Chinese were individually divided into groups by CURE clustering technique and the suitable learning resources were recommended to the corresponding learners. Thereby, the blind problems of network autonomous learning on choosing resources were solved.