Research On Personalized Recommendation System In E-learning

With the rapid development of World Wide Web(WWW) technology,Internet has become one of the most important information and knowledge sources.E-learning has achieved great success both in research and applications in recent years,and provided learners with unprecedented abundance of learning materials and great flexible learning ways.However,because of several reasons,most E-learning system still focus on the web site only,using just-put-it-on-the-web approach,which leads to the case that learners need to spend a lot of time and effort to find resources that they enjoyed and required.For this reason,learners hope the system to provide personalized service based on their different characteristics.Therefore,to design and to build a personalized E-learning environment has gradually become the focus of many researches.Applying data mining and agent technology to E-learning environment and building a personalized recommendation service is a hot topic in this area.Currently,there exist many researches about personalization in E-learning and made some progresses.Personalization platform has already appeared in practical application.But because of the complication of E-learning environment,reasonable and timely understanding and expressing learners' needs,interests and hobbies is a fundamental and difficult problem.The current widely used methods heavily rely on content analysis of resources,which itself if a difficult problem.In addition,the use of efficient mining algorithm is also providing strong effectiveness of the recommended result.The main content of this thesis includes:(1) combined with the specificity of personalized service in E-learning,this thesis introduced the research background and significance of the needs for personalized service,and summarized the recommendation mechanism applied in E-learning.(2) in order to represent and analysis of learners' interests and hobbies,this thesis proposed a method which is used to divide learners' historical session based on a special time interval,and given a user similarity measure method based on this dividing method.(3) this thesis proposed a new clustering method based on ant colony and genetic algorithm,and a detailed description of the algorithm and experimental validation.(4) this thesis build a personalized recommendation system model based on web mining and agent technology.This mode has 2 layers,and each layer is composed of several special Agents.Then,the system model is validated by experiment,and this thesis also analysis the factors affecting the recommended result.This is all the bases for practical use of the recommendation system.Finally,we concluded and summarized the study,and put forward the direction of further research.