Ontology-based Search Engine Under P2P Environment

The network epoch is coming with the rapid development of Internet throughout the world, and with the high-speed evolution of the Internet and the World Wild Web (WWW) technology, information dwelling in the Web explodes dramatically. This makes it difficult for traditional search engines, which commonly run on central servers, to search desired information efficiently, accurately, and comprehensively. Therefore, the search engine technology needs further investigation to meet the change of the new situation.In this thesis, we introduce a peer-to-peer (P2P) distributed searching model to extend the scale of searching engine and to improve its performance as well. Firstly, we investigate the key technologies, developing trend and challenges of the traditional search engine technology. Secondly, a new method of combining the advanced P2P methodology and the traditional search engine technology is presented, which fulfills the needs of distributed and localized search engines. The advantages of P2P distributed searching model, such as high performance, deep and wide scope in content, high precision in searching result, are analyzed.In order to resolve the compatibility problem existing in the P2P systems, JXTA is adopted as a general P2P development platform in this thesis. Based on JXTA, a basic P2P network is constructed, a specification of P2P search engine is defined and implemented, and a localized search engine model with high scalability is presented. This offers a feasible solution to improve the scale and capability of search engine system.The study of P2P search engine is still in the theoretical phase. This thesis gives a methodology to construct the search engine system based on P2P distributed searching model, which dedicates to the design and development of a valuable search engine system in the future.