Research Of Intelligent Search Engine Based On Semantic Web

With the development of the computer technology, Internet has become popularity and being an important way to obtain knowledge and resource. People could look for information by search engine on the internet, but the traditional search engine is hard to find the information what the users really want because it is short at unified resource description, so the intelligent and semantic exploration is an important discussion to the research of search engine.This thesis analysizes development of intelligent search engine, and studies interrelated technology of Semantic Web and ontology. Based on the deep analysis and research to the traditional search engine and semantic search engine, the thesis syncretizes the technology of information retrieval and the character of Semantic Web and ontoloty knowledge, and it finally proposed a model of semantic search engine. This search engine model is divided into three design points, namely the design of ontology database, the semantic reasoning technology and the ontology matching algorithm. Semantic retrieval need formal and semantic information, and it extend the information. The knowledge information is described formally with ontology database, and semantic extendibility is implemented with an ontology matching algorithm and semantic reasoning technology. Concepts-weights vectors team matching algorithm is an excellent ontology matching algorithm, but its application scope is too small, in the case of concept unshared, its precision declined. The thesis improves on the algorithm and an ontology matching algorithm based on semantic similarity is proposed. The new algorithm expands application scope and improves the accuracy.Finally the proposed model is verified through three experiments. Semantic reasoning is verified with the standard domain ontology. In the second experiment, the proposed semantic similarity algorithm proved an effective algorithm. The last experiment verifid an ontology matching algorithm based on semantic similarity is better.