Research And Application On Semantic Retrieval For Digital Library Based On Ontology

With the development of knowledge economy, the digital library as an important carrier of the spread of information resources, the current digital library retrieval system searching based on keyword, and the only match by characters, It can not express semantic information, so it can be produced bias between search results and user's demand, resulting in retrieval recall and precision was low.Based on the traditional digital library search model, this thesis uses ontology of semantic technology to make a in-depth research into retrieval model, and puts forward a digital library search model on the base of ontology structure. This model which is based on domain ontology, composite technology of extension search, semantic reasoning, semantic indexing, and resources mapping completes semantic retrieval.This thesis builds an ontology based on the model. Take computer books as example. It collects information from fields, analyze and classify them, and build a complete subset of computer books.The digital library retrieval model is implemented on the base of semantic ontology. According of the relationship among the concepts of ontology, the resource mapping model is designed, matching the concepts and properties of ontology library to database table pattern is implemented, and the mapping is built between the database and ontology. OWL ontology language is used to describe the ontology by Jena2, ontology is stored to database and semantic is reasoned by Jena2. Based on the model a prototype system is designed, including ontology database maintenance module, resource mapping module, searching module and output module search results. Semantic structure model of the digital library in the recall rate and precision is superior to the traditional structure by experimental data analysis.