Ontology-based Vertical Search System Design And Implementation

Accompany with the explosive development of information technology, network becomes the primary way of obtaining information for people. Network endows convenience on people but puzzle people on the other edge. As information is explosive, enormous and unsystematic, finding useful information from the Internet always waste people a lot of time. As a result, information searching technology already turned into a necessary tool. As the prevalence of usage, information inquiring technology is pushed to fast evolution. The performance of search engine poses a big influence on the speed of catching the effective information in the Internet.Massive information on the Internet causes 'information overload', how to search information in a quick and exact way is becoming an emergency problem. Aiming at given domain or industry, vertical searching engine made a professional and deep analyze, classify in a refinement rule, and filtrate in a certain rule, it can get rid of information weed which is not concerned with search contents, as a result, it prove that it possess higher veracity ratio and better information pinpoint capability. It can be concluded that vertical search is better and stronger in special domain search, and it will be more popular than other search engine.After analyzing the current and deficiency of the domestic and abroad search engine, basic knowledge of semantic web and ontology theory, and apply instance of them, the paper proposes an information search frame base on ontology at the actual apply aspect: snatching relative network page in a more effective way according to set the theme of searching content, pumping out useful information after trimming un-relative information, and then exhibiting those selected information to user. The key point of the paper is the information-extraction module design of topic-oriented search engine system by adopting certain algorithm, and this module is composed by page obtaining, page pre-handling, topic information exaction and inquired content matching. The actual usage proves that the otology search engine with information-extraction designed by the paper improves the veracity and speed of information obtaining enormously.Using JAVA environment to construct the system frame, MVC and Struts to develop, protege ontology modeling tool to build ontology, XML language to describe ontology, the system started from actual apply angle, classified the search contents topic, snatched the relative pages from the web in a more effective way, then after scratching up useful information, and finally showed the search results to the user in a certain way.