The Design Of The Auto Industry Vertical Searching System Prototype And The Implementation Of The Primary Module

With the development of internet, researchers pay more and more attention on Information Extraction (IE). Information Extraction can extract latent and valuable information from internet. These valuable information can be used into management and decision. Search engine resolved some difficulties for users. However, common search engine have some default, such as how precision ration and simple function. So vertical search engine can overcome these difficult. Our aim is to build a vertical search engine.In our dissertation, we explore the theory of vertical search engine, and introduce the rule of web spider. Information Extraction (IE), and the technology of vertical web site. Chinese gazetteers and Chinese named entity recognition are difficulties in IE, and we propose a way to resolve these difficulty. We build gazetteers that is made by GATE about automobile domain and establish rule of JAPE to a Chinese named entity recognition. At last, we build a vertical search engine about automobile.At last, We build a auto mobile vertical search engine which is based on Nutch and Gate. The architecture of the automobile vertical search engine include three part, spider, Chinese IE and Chinese information retrieval. Our system achieves vertical search in auto mobile, and can gat good results.