Research On Subject Information Collection Based On Ontology Evolution

Since the invention of the Internet, people have been using it as a new way to obtain information.While people own an exploding information source, they have to face with problem that how to findspecific subject related information from massive amounts of information from the Internet. Generalsearch engine is the most popular information retrieval tools, but it could not provide timely andaccurate information of specific subject as it is for the mass. In this situation, subject orientedinformation collection has been a research hotspot.This paper first shows a brief overview of researches at home and abroad on subject informationcollection technologies and ontology evolution, presents fundamental, architecture and developmentof web information collection, introduces theories about text categorization and ontology. Secondly,put forward collection strategies for several information source on the internet, including wholewebsite crawling, target block tracking, RSS incremental updating. Then, design a solution forontology evolution which contains webpage content parsing, keyword extracting, original ontologybuilding and ontology evolution. At last, establish test system and verify the solution with samplesubject and original subject ontology.The main work of this paper consists:①design the collection strategies for various informationsources so that information-collector can adapt to the complex environment of Internet and obtainsubject-information from several information source under guidance of subject-ontology;②putforward a semi-automatic ontology-evolution solution which is based on web-pages and user behaviorlog, combined with keywords extracting technology, and is aimed at achieving subject-ontologyevolution under users’ guidance. And an experiment is designed to verify the validity of this solution.