Semantic Web-based Information Retrieval System

The explosion of information is taking place as the widely uses of Internet application ,which makes people get information they wanted by every way, up to now, many tools of IR(information retrieval) and search engines have been used by people to get knowledge and information in the life and work. Old search engines are only based on the match of word and expression on the level of programmer, and they are lack of ability to represent,process and understand knowledge. How to get perfect and precise information effectively has been a puzzle for people. This fact is studied in this paper.This paper firstly introduces research background of IR, analyses status of IR and search engines at present, deeply researches the technology of IR based on semantic web and upgrades IR from grammar match of key word to semantic match of knowledge(or context),in order to enhance recall and precision of retrieval, which is aim at the disadvantage of IR and deficiency of search engines at present. The main study as follows: the conception and technology of ontology and semantic web, and specific application on IR; the search technology of intelligent information, including of the method and technology of building repository based on ontology by the tool Protégé, and data query and simple knowledge reasoning based on Jena; the method of describing domain ontology and auto semantic annotation of text information resource based on electronic consumer goods, building simulation repository, searching information and extracting knowledge content in related area.The innovation of this paper is presenting a improved method of auto semantic annotation based on text resource of domain ontology. This paper researches information retrieval based on semantic web and builds system model of SIREP(Semantic Information Retrieve of Electronic Product),adopts method of auto semantic annotation for intelligent annotation by machine learning. The aim of research in this paper is to enhance recall and precision of IR, remove the redundant and irrelative information, realize effective mechanisms of satisfied retrieval for users.