The Enhancement And Application Of Collaborative Filtering Recommendation Algorithm In E-learning System Based On EGL

With the continuous development of network and information technology, e-Learning which is a way of online learning based on Internet, has been gradually played an important role in the information construction and talent cultivation in enterprises. However, some enterprises such as the large number of small and medium enterprises can not provide enough capital and technology support for their e-Learning construction because of the scale limited. So the application architecture must be proposed and implemented to provide rich human-computer interaction experience for learners, but also can reduce the development cycle and cost of e-Learning system.With the explosive growth of Internet information resource, majority of current e-Learning system also has some problems in the practical application. Because the e-Learning system does not fully equipped with the features of personalization etc, learners often can not find suitable curriculum resources to their own conditions and preference. These are not conducive to improve the learning efficiency. Therefore, to provide personalized resources service for learners is an important part that e-Learning system should be enhanced at present.The research in the paper aimed at designing and implementing an e-Learning system based on EGL to create a good online learning environment, provide personalized knowledge resources navigation, and reduce the complexity and the development cost of e-Learning system in a certain extent.The main work in this paper includes:(1) Aimed at the problem of high development cycle and cost what developing e-Learning system using Ajax framework brings which has high complex of technology foundation, the research presented a hierarchical scalable e-Learning system using the features of across platform and across application of EGL on the basis of analysis of the relationship between EGL and Web2.0. The layers can cooperate with each other to complete the task, but also can be relative independently, so that developers can focus on the business problems what the code handle rather than technical details. The application architecture can not only simplify the development process of e-Learning system and effectively reduce the development cycle and cost of the e-Learning system, but also improve the human-computer interaction experience and real-time requirements.(2) Designed and implemented a personalized e-Learning system based on Collaborative Filtering using the above-mentioned e-Learning system application architecture. To predict the knowledge resource what the target user may be interested in or need through the existing users’ past behaviors and preferences in the system, and thus as personalized knowledge resource navigation for learners. The personalized e-Learning system can recommend the learners individual lesson items that are appropriate to them and help them to avoid trapping into the trouble of knowledge resource disorientation, thus help them to improve learning efficiency.(3) Aimed at the issue of new users’ Cold Start of Collaborative Filtering recommendation algorithm, this paper introduced several concepts and methods, such as similarity matrix etc, to proposed an enhanced composite recommendation algorithm based on Collaborative Filtering and content-based tags extracting, and then conducted experiment using the new enhanced recommendation algorithm with other three traditional common solutions to Cold Start problem. The analysis of experimental data verified the accuracy and effectiveness of the algorithm. The new users can acquire appropriate curriculum resources and enjoy the personalized knowledge resource service.Three EI papers related to this research have been published by the famous international conferences and journals.