A personalized recommendation system in E-Learning environment based on semantic analysis

The proposal of various electronic learning contents, e.g. remote education or virtual classrooms, has given a powerful impetus to the E-Learning techniques. However, there still remain several hard and complicated problems unsolved. Especially when compared with e-commerce and medicine, the problems of the recommenders in E-Learning system have not been fully figured out. In this paper, a novel personalized semantic recommendation system (PSRS) for E-Learning is designed. The proposed PSRS system employs the Video Structurized Description (VSD) technique to extract the initial keywords description of the learning contents, and then adopts the lexical parsing technique to refine the descriptive words with a standard format according to the initial keywords. Subsequently, the PSRS adopts rules auto-updating (RAU) to automatically add sequential items into ontology rules. Depending on the specific ontology knowledge with domain rules, semantic mapping and intelligent reasoning techniques are applied to generate certain semantic related recommending items for the active learners. Experimental results indicate that the proposed PSRS preforms better in accuracy than any other existing algorithms.