Research And Implementation Of Semantic Vertical Search With Distributed Data Sources

In recent years,with the rapid development of Internet technology and theexplosive growth of web information, the Internet has become a massive data space.It becomes more and more difficult for people to get the information they need fromthese massive data.So search technology has become particularly important. Withthe rapid development of search engine technology,general search engines arebecoming increasingly powerful functionality. This makes the structure of generalsearch engines increasingly complex. Beside this, it has a huge index data and verybroad topic and no classification between themes and topics. Not enough in depth,lower precision and timeliness of general search determine these limitat ions such aspoor general search does not meet specific user needs. There is an urgent need to bea more targeted, more depth search engine for individual needs of specific industriesand specific areas. In this case, the domain-specific vertical search technology cameinto being. At present, there are some vertical search engine, such as Kuxun, Qunarand so on. But these vertical search engines still use keyword-based search methodsand can not meet the user-level semantic retrieval requirements.To study and implement semantic vertical search, to improve the quality ofretrieval, this paper studies the semantic vertical search technology in the distributedenvironment, including construction of the domain ontology, semantic queryexpansion model based on domain ontology, information extraction based on DOM,using Lucene4.x to index data, and at last making experiment analysis. This paperstudies the following contents and key technologies:(1) Constructing domain ontology in field of computer security by collectinginformation, filtering, classifying. And make description of defination for classes,properties, relations between classes.(2) Extracting the information of computer security based on DOM rules.(3)Indexing data of classes, properties and relations between classes.Complishing user queries and information retrieval.(4) Constructing word hierarchy based on the provided relationship category.At last, this paper implements semantic vertical research based on domainontology using RESTful ideas. Experiment shows that Computer Ontoloty Searchresolves the problem existing in traditional search based on keywords. And It showsthat semantic vertical search has practical significance.