Concept Based Learning Contents Retrieval by Using Extended Vector Space Model with Ontology

For efficient learning procedures, it is important to provide the learners with contents that are appropriate for their intentions. Existing contents searching systems used statistical methods to estimate the meanings of the contents, or expansion of user query to find the contents that the learner wants. However, these existing methods failed to efficiently convey the intentions that the user wants, since the methods do not identify the topics directly from the learning contents. In this paper, we suggest an algorithm to identify the context of contents using domain ontology. The algorithm takes variables of sub-super concept relations of the domain ontology and relation information of properties between concepts to identify the topics. Also the proof of the superiority of the algorithm compared to the conventional keyword-based method was provided through constructing a domain ontology related to middle school mathematics, and experimenting with one thousand contents.