The Information Extraction Technology Research On Open Knowledge Base For Entity-oriented Query

Query recommendation as an important technology used in search engines suggest s relevant queries to help users to reformulate more accurate queries. Existing approaches of query suggestion compute query similarity based on direct matching of query properties. However, it is hard to find the semantic relevant queries that are related indirectly.To solve the above problems, this paper adopts Wikipedia as the open Knowledge Base and proposed a new type of query expansion system. This method by extracting information from the structural part and natural text in Wikipedia,has formed a semantic systematic corpus in which the entity as skeleton, physical characteristics and entity relationship as the network, based on this corpus complete corresponding user inquires recommendation system, and designing auxiliary query system to improve inquires recommended effect while the user inquires won’t collected in wikipedia.The main innovation of this paper is as follows:1. This paper proposes a novel methodology to the problem of query intent classification based on random walk model,which is called RWM. This methodology deals with the problem to sparse data, according to random walk model processing, aim to get connection with indirect reletavent contents., consequently result show the method is provides an effective solution to query intent classification.2. This paper proposes a classification algorithm for rare query using Wikipedia and web knowledge, which is called WWRQ. we determine its topic by classifying the web search results voting whit information properties from Wikipedia.The experimental results show RWM offers both more accuracy and recall rate compared traditional inquires system, and WWRQ efficiently solve problem that short inquires can’t be accurately orientation.