| The data on current Web can not be understood by the machinebecause these data are user-oriented. Therefore, the data retrieved throughtraditional retrieval means can only show one facet of the contentrepresented by data. When computer can not precisely understand user'srequirements and Web data, semantic obstacles between Web data andusers arise. As a development of current Web, Semantic Web aims toenrich the Web data with semantic format and make themmachine-readable. Thus it will be convenient for human to co-operate.The thesis achieves next results:Firstly, semantic Web proposed by Tim Berners-Lee and topic-basedinformation retrieval have been introduced. Based on analyzing currentsituation of semantic Web and topic-based information retrieval research,the research direction in the thesis are put forward.Secondly, ontology construction, Web data collecting, Web pageanalysis, topic relevancy and etc. are elaborated. The feasible solutionshave been respectively proposed for each issue discussed above. Theseprovide theoretical and practical foundation for upcoming design ofontology-based Web information collecting system.Thirdly, Ontology-based focused crawler system (Ontowing) isdesigned and implemented. The framework, working procedure,components and functionalities of Ontowing are elaborated. As asub-system of SNAX, through combining semantic Web technology withinformation retrieval technology, Ontowing realizes user's relatedinformation and resources collecting.Finally, experiment has been done to justify the proposed theory.The summary and expectation of the research have been made. |
PDF Full Text Request