Research On Recommendation Strategy Of E-Learning Personalized Recommendation Systems-Revelations From Commerce Recommendation Systems

In the background of digital lifelong learning system construction, E-learning personalized recommendation system,which is one of the most important systems of lifelong Learning, has been widespread concerned. Personalized recommendation system (hereinafter referred to as PRS) applied to the fields of e-commerce and information services early, and has been relative mature. But in the E-Learning field, PRS application has been in the stage of researching. In view of this, the author will take e-commerce PRS as the breakthrough point, study the successful case in e-commerce PRS, and get some revelation from their successful experiences of application, finally explore the recommendation strategy of E-Learning PRS.In this research, the theory of personalized recommendation is theoretical basis. First, by means of documentation, this paper discusses the connotation of the personalized recommendation theory, then briefly introduces the key recommendation technologies, compares with the advantages and disadvantages of those technologies, and puts forward the personalized recommendation strategies.Second, the case study method, with amazon.com, MovieLens.org and douban.com for case study, this paper analyses their typical functions, recommended strategies and advantages, and then get their successful experience.Third, it analyses the form of E-Learning PRS. Compared with the similarities of e-commerce recommendation systems, the following aspects can be used in E-Learning PRS: 1. establish virtual Learning community; 2. use tag and correct tag; 3. fully explore the recommendations among users; 4. improve recommendations, 5. establish personalized learning environment.Finally, it puts forward the personalized recommendation strategies of E-Learning PRS: using the combined recommendation strategy of collaborative filtering technology and association– based technology, establish virtual Learning community. And then using the combined method of system recommendation and users recommendation, recommend to learners a complete study plan, including learning resources, learning activities and learning strategies.