Research On OWL-Based Semantic Query Technology

The traditional information search technologies are mainly keyword-based match.Its query performance is not as good as required.Users have to spend a lot of time to search in the query results,which usually can't find the information they need. Once 'Semantic Web' was proposed by Tim Bemers-Lee,its theories and technologies have become a hot research area.OWL has been recommended by W3C as a standard web ontology language and it has a good development prospects. This thesis focuses on OWL-based semantic query technology.First of all,this thesis does some basic research on the architecture and key technologies of semantic web,ontology,semantic query and its method.It also analyzes OWL systematically and classifies the language norms of OWL in detail. Then,based on the investigation on semantic query technology,an OWL-based semantic query model is designed.Then following its work process,each module is carefully explained one by one.The main research work in this thesis is listed as follows:(1) The ability which supports the semantic match among different languages and query the documents of different languages are integrated into the query interface of the model.(2) For the two given semantic reasoning extension rules of 'is-a' relation and 'equal-to' relation,it discusses the design of the algorithm,including the choice and design of the data structure.(3) For the four different query objects RDB,semantic DB,web pages with semantic annotation and web pages without semantic annotation,it discusses their semantic query methods respectively.For the web pages with semantic annotation (based on OWL ontology),a semantic query method was designed.(4) By doing some research,an existed calculation method of semantic relevance was supplemented,and a calculation method of semantic relevance was given.Using this method,we can calculate the semantic relevance of the query results in different ontology.By which,we can sort the query results and control the number of returned results.(5) Let the traditional query engine work independently with the semantic query engine in the model,which can make up the premature of semantic query technology. The ability which supports the choice of query engine is integrated into the query interface,which can control the two query engine's work.At last,it analyzes the feasibility of the model from its implementation technologies and query performance.