Research Of Natural Language Search System Based On P2P Network

Nowadays, the P2P technology has become the focus of research. Meanwhile new applications of P2P emerge quickly. Under such a circumstance, this thesis mainly discusses how to find the data better via P2P platform for users.Now, we are all used to finding what we want through the network, but this kind of search is based on the servers of the network. As we know,the search speed of the search engine based on the server is high,but the search ratio is very low: It can only find the web files. On the contrary, the P2P search engine can get almost all the content on the network and highly improves the utilization ratio. So in this thesis we explore a way to search directly through the P2P network and let the end users get what they want by some natural words rather than the key words.In order to realize this idea,we integrate a natural language word-cut module with a special designed word dictionary into the P2P system. Further more, in order to decrease the diffuse path number, we designed a new P2P topology with the reference to the semi-distributed network platform. Also we propose the algorithm. At the end, we test our idea through a simple system.