Research Of Semantic Retrieval On Instructional Resources Based On Ontology

The inflation of network data and the rapid development of information technology bring huge opportunities and challenges to the education industry. As the network electronic Learning (E-Learning) becomes a new study method, in the opening distributed network environment, the mass of learning resources description, management and retrieval becomes research hotspot. But, because that there is no uniform standard for education resource description and description structure of subject knowledge causes resources discrete and difficult to share and cannot understand resource, knowledge and relations among them semantically. In addition, because of the lack of semantically understanding of key words in traditional full text search mode based on key words, the retrieval results often contain much irrelevant information, and the search mode cannot establish query on the non-text learning resources.Semantic and ontology has become a new method for knowledge and data representation since the proposing of semantic web. According to the requirements and some existing problems in E-Learning applications, it is suitable to use ontology technology for subject knowledge representing and instructional resource description, and to carry out retrieval learning resources through a semantic way. This is also an important development direction for E-Learning, theoretically and practically.In this paper, we studied the semantic Web, ontology, semantic query expansion and semantic data retrieval technology, and proposed the knowledge hierarchical model with knowledge point as the unit through careful analysis on subject knowledge description, the knowledge structure and the knowledge relationship. The domain knowledge ontology for the course data structure has been constructed based on the model to scientifically organize the subject domain knowledge. In reference to the domestic and foreign relevant learning resources metadata standards, though the construction of instructional resource ontology on the basis of CELTS learning object metadata standard, the connotation and extension of the learning objects and the relationship with subject knowledge units has been discovered. In semantic web, the query extension method based on ontology has been proposed. Firstly, this method extracts search terms from the query information, analyzes and calculates the relationship between concepts. Then a search tree has been built to express the users' search intents. According to the relationship between subject knowledge unit the search tree has been expanded semantically to enrich users'query messages. In processing of instructional resource retrieval, index weight has been determined for different instructional resource metadata, and the semantic vector has been established on the metadata to express the importance on the resource content it reflects. By calculating the similarity of query-resource, the initial query result has been obtained. Then other related resources are retrieved through instructional resource ontology reasoning on the initial query result. Finally the retrieval result has been sent back to user.This paper finally implemented a semantic retrieval model system of instructional resources based on ontology to validate the semantic query method. Through the experiment contrast and data analysis, the semantic retrieval method in this paper can get better retrieval quality in learning resources of various formats, performing higher in recall and precision than traditional keyword based search algorithm, reflecting the advantage of the method.