Research And Implementation Of Personalized Teaching System Based On Learning Behavior Analysis On E-learning Platform

The rapid development of information technology and network is changing people's learning behavior and communication behavior. More and more learners take part in network learning. E-learning platform is an important approach to network learning, it provides learners a rich learning resources and convenient learning services. However, many E-learning platforms are still materials-centered, which results in that learners facing a massive of data in E-learning platform often show confusion and lose the initiative in learning process. Moreover, the personalized demand of the learners with different learning pace, different learning methods and different interested can not be fully met. For solving the problem in network learning, in this thesis, a learner-centered learning behavior analysis model is built based on an old E-learning platform for analyzing learners'learning needs, learning interests, and learning behaviors, and providing learners personalized educational resources based on the learning behavior analysis model.In the process of constructing learning behavior analysis model, the following contents are studied in this thesis:1) By behavior sciences as a basis, based on the "people-centered" principle, through the survey, a detailed analysis of the characteristics of learners, learning resources, and learning tools is done for finding learners'usage demand in E-learning platform; 2) Using web usage mining technique, the information of learners'behaviors is collected and collated; 3) For improving traditional association rules mining algorithm which has frequent sequence set in the forms of organization, a conjectural set between questionable set and believable set is added to improve the speed of finding association rules; 4) Based on object-oriented methods, the stored forms of learning resources are reconstruct. According to package size of learning resources, the atom-study objects and multi-study objects are reconstructed to enhance the utilization of learning resources.The model and system proposed by this thesis can organize learning resources reasonably and meet the basic learning purpose of learners'in network learning, track learners'behavior, find learners'habit, analyze and summarize learners'learning rules. Based on the results, reasonable learning resources and learning paths are recommended to learners for more intelligent and personalized E-learning.