Personalized Learning Resources Semantic Retrieval Based On Ontology

With the massive growth of learning resources, since different users have different background and different knowledge, the interests of the users are different and so needs of users are different in information. Traditional information retrieval based on keyword matching technology in resources coverage and retrieval precision are unable to meet the needs of personalized learning behavior and creative learning activities. Therefore, how to obtain useful learning resources from mass of data and information efficiently to meet needs of the various different information of different users become an urgent problem.In this paper, based on a systematic study of semantic retrieval and personalized learning, a personalized learning resource retrieval models based on semantic have been proposed. Models construct domain ontology of learning resources to achieve the process of semantics, and build user interest- models to capture user interests to accomplish personalized learning resources semantic retrieval.The main contributions of this paper are listed as below:1. Introduced ontology modeling rules and procedures briefly, analysised the relationship between metadata and ontology. Reference to seven-step method, Ontology was constructed with the combination of CELTSC educational resources basic norms and the relationship of knowledge point, and explained its feasibility through the example.2. Constructed a personalized user interest model framework. At the same time, the user interest drifting problem is studied. Making use of the concept correlation matrix to calculate semantic associations for merging a hew interest and the current interest between concepts between the user and the current interest so as to achieve an update of user interest.3. Studied the similarity computation, the retrieval algorithm which the semantic retrieval process involves and so on the key technologies. Carried on the experiment to test the algorithm validity, and confirmed the algorithm feasibility effectively.4. Proposed a personalized learning resources semantics retrieval model based on ontology. The model is constructed base on learning resources library and user interest model construction. Through calculate the concepts similarities between learning resources to realize personalized learning resources semantics retrieval.This next phase of work will focus on improvements of user interest model algorithm,so as to provide users with more accurate and personalized services and improve the performance of e-Learning Platform.