Study On Information Retrieval Model For Digital Library Based On Semantic Web

With a rapid development of network technology in recent years, network has now become increasingly ubiquitous. The information share which is provided by digital libraries has been more and more important. At the same time, new challenges have been brought in by the development of Internet. The problems of traditional information retrieval for digital libraries have emerged. For instance, the data are distributed, isomerous and the systems can not understand the real retrieval intentions of users, so that low retrieval effectiveness is caused. Required information is omitted, or the information which is not required is provided. The users are not satisfied with these. The emergence of semantic web technologies can make the performance of information retrieval become perfect. Therefore, by using semantic web theories and the situations of the digital library information retrieval, an intensive study of digital library information retrieval based on semantic web is made in this thesis.Firstly, the concepts and technologies of digital library information retrieval and semantic web are introduced,the problems of digital library information retrieval are analyzed also. On basis of existing digital library information retrieval model, a new one based on semantic web is designed. Three important modules and their functions are discussed in detail. Its system structure is discussed either. Appling the technologies of semantic web to digital library information retrieval, the model makes use of the semantic relations among ontology concepts in the framework model to direct the semantic retrieval process of digital library. Thus, this model can perform the information retrieval and inference based on the semantic relationship in the documents, which enable the users to find their resource more rapidly and precisely. The recall and precision of digital library information retrieval can be increased.Meanwhile, an in-depth analysis and study of key issues in information retrieval for digital library based on semantic web are also demonstrated in this thesis. The methods of domain ontology integration and construction are analyzed in detail. Besides that, the three different methods of semantic similarity computation are researched. In the semantic distance approach, semantic depth and the degree of ontology are considered also. It is improved. The conception features approach is improved also. The information measure approach is taken into account. On the basis of these, a corresponding semantic similarity computation method is proposed finally. In this part, the study of the key issues provides theoretical support for the research and perfection of digital library information retrieval based on semantic web.