Research On Personal Learning System Based On Data Mining

Currently, learning based on Web has made considerable progresses under the guiding of education theory, but there are still some problems existed, such as singularity teaching mode, educational resource stacking simply, low intelligentized level etc. But different student has different teaming capability, personal interest, personal learning foundation and struggling level. Contradiction grows up between distance education and personal demand of student. Therefore, these issues should be studied and resolved further.Thorough researches are made to understand the characteristic of learner access patterns by constructing the personal learning system. Major works are listed as follows:The research on E-learning behavior and Web data mining, which is the key technology for designing personal learning system , is carried out.Analysis the algorithm based on Frequent Link and Access Tree (FLaAT),which is proposed to mine frequent user access patterns. FLaAT stores all user access information and considers the merge of same routes with different prefixes to make the integration of mined information during search of frequent user access patterns.An efficient algorithm is provided to mine learner preferred access patterns based on the concept, which is along with support to disclose learner interest and preference. It considers all possible factors, weight of web page and relative access frequency of web page. Thus it more deeply reveals learner interest and preference. And a fuzzy approach is proposed to cluster learner access patterns from web logs.The paper constructs the application model of data mining in the personal learning system, and discusses the personal learning model and data ming model. Finally, we make a conclusion of the paper and point out what to do in the future.