Research On User Interest Modle And Application Based On Learning Behaviors

With the increasing popularity of E-Learning’s application in the field of education and business world, providing individually-oriented learning service, which targets at user’s interest and needs, has become a main direction in the development of E-Learning due to the fact of huge differentiation of each individual learner with the aspect of learner’s ability, foundation and attitude. The construction of user’s interest model is the core and key to realize individual-oriented learning service. However, the majority of current user interest model constructed on the base of the content that user browsed or the mining of web log. These methods can get user’s interest while not complete and accurate, which results the poor quality of service that provide for users.This paper proposes the method to construct the user interest model based on users’learning behavior. User’s learning behavior can reflect user’s psychology activity with individual differentiation among different learners, which is inevitably caused by the learning process itself. As a result, gaining users’interest by their learning behavior in order to provide individual-oriented service will be more accurate and targeted. The paper analyzes these aspects based on the above introduction:(1) Analyze the user’s learning behavior under the E-Learning environment and introduce the management of knowledge points and the classification of learning behavior. Provide the concept formalization of each learning behavior.(2) Construct the class resource by knowledge unit and knowledge point. Introduce ontology concept for knowledge and class in order to form a complete logical construction of the whole knowledge system.(3) Quantify user’s interest rate by Multiple linear regression model. Compared with other research result that only indicate user’s interest on certain web page without accurate quantification, the method proposed by this paper can calculate the exact number of user’s interest, which is also can be validated by the Significance analysis and test.(4) Indicate the model of user’s interest through the analysis of page-interest and other weights in knowledge system.. Provide recommended individual-oriented service by user active recommendation algorithm and user passive recommendation algorithm on both user’s interest orientation level and cognition level.