Semantic-based Learning Resource Difficulty Assessment And Recommendation

In order to make full use of online learning resources and help learners choose the bestresource they need, this paper studied on the methods of assessing the learning resource difficultyand personalized recommendation.Firstly, the paper establishes a learning resource difficulty evaluation and recommendationframework, which contains resource layer, syntax layer, semantic layer, service layer andapplication layer, and the related technology is described.Secondly, a Chinese computer science domain ontology is automatically built based onWikipedia classification data, which provides the semantic basis of calculating semantic similarityand knowledge point difficulty.Thirdly, the study analysis the factors of learning resources difficulty from three main aspects:content difficulty, organization difficulty and expression difficulty, and presents a learningresource difficulty evaluation method based on ontology and Analytic Hierarchy Process. Themethod helps learners to choose the learning resource with suitable difficulty by providingdifficulty score for reference.Fourthly, the paper presents a document vector model contains both difficulty and knowledgepoints, through which a user demand oriented learning resource recommendation method isestablished. In addition, a query interface is implemented based on an open source search enginewhich supports keyword and document querying.Finally, the implementation of the application is introduced, including ontology-basedsimilarity calculation, document data extraction, semantic retrieving and recommendation.