An Collaborative Filtering-Based Personal Recommendation System Design And Implementation

With the development of Internet technology,traditional education industry is starting to suffer from the impact of the Internet, cloud education and online education has gradually become a research hotspot.due to the current form of online education are unitary,and unable to analyze the problem of students in learning,which decrease students' learning efficiency.The student can not find suitable learning resources,when they are facing massive resources, as a result the online education can not give full play to intelligent tutorin.In this paper, by constructing the ontology knowledge in the field of elementarymathematics, establishing users and the concept of knowledge model, using collaborative recommendation algorithm to produce knowledge andrecommend topics.The details of the research are described as follows:(1) This thesis researches the theories of ontology, mainly includes the process of building ontology, and then using the mathematical process of building domain ontology to describe that how to define classes in the ontology knowledge hierarchy, how to define class attributes and constraints,and how to create an instance of a class.(2) This thesis researches on ontology concept of correlation degree computation method of knowledge.Through the analysis of the Semantic relevancy of four factors which are semantic distance, semantic overlap,the concept of level difference, semantic weights, and the extendtion of the concept relations collection in relations of two concepts, we can achieve the concept of ontology-based semantic similarity calculation formul.(3) This thesis use traditional collaborative filtering algorithm to get concept and topic recommendation information, based on ontology and concept of correlation degree.In this algorithm we created a user-concept model with the learn history, extended the user model matrix with concept of correlation degree,and calculated the nearest K users, finally we get the recommendation information.(4) The system of personal recommendation come true with all above theories and technology. It can provide users with an autonomous learning platform,record and analyze user learning information,the system will recommend appropriate learning resources to users by using collaborative filtering algorithm.