Research And Application Of Vertical Search Engine In Manufaturing Resources

As a part of the project of "the key technology of network manufacturing resource management aiming at Cooperation business" and "The ASP manufacturing platform aiming at machine and electronics" that undertaken by professor QS.Xie, the research project was financed by "The National science finance "and "The Main Plans of The National Project 863 / CMS ".The effective and accurate retrieval of manufacturing resource is the base of enterprise informatization, it is one of requisitions to realize the network manufacturing as well. The key is how to select the appropriate manufacture resource combination from dispersive manufacturing resources then satisfies the demand of fast processing, promoting the new product. This paper pays an emphasis on studying the technologies of the system, including the topic crawler and the Chinese word segmentation. The technologies of the topic crawler including how to obtain the seed links, the cravvle algorithm, the analyses of the similitude degree, html parser, the technologies of abstract and so on. To improve the efficiency of the crawler, a model of page evaluating was added into the crawler module; therefore the urls in a page with a high similarity of the topic will be first crawled. A improved word segmentation was proposed by redesigning the store formatter of the major dictionary and the algorithm of word segmentation on the Chinese word segmentation module.The lucene technology was used on the module of the index and retrieves of the system. To improve the efficiency of the lucene's word segmentation and meet the system's requirement of world segmentation, a new word segmentation algorithm was added to instead of lucene's. The resources code was written in Java, a object oriented program language. To make full use of the computer resources, the technology of muti-thread was used in this system. To get the concise web pages and make different pages of showing the results and producing the results, the server script language JSP was used, with the same advantage of the Java. A test was carried on at the end.