The Research On Ontology-Based User Interface Method Of Data Integration

The environment of data storage becomes more and more complicated and the style of information representation is also diversified wherever was in web condition or local organization. How to construct the unified Data Integration System (DIS) to support the user's decision and supply consistent and lucid information services for users by using the distributed and heterogeneous data resource has become a new problem for the development of information technology.Based on the research of traditional user interface method of data integration and the project of our laboratory that is DIS based on Data Service Matching, this paper proposes an idea that UI are composed by graphic interface, consistent data view and processing of users'query, designs a relatively natural and intelligent solution of UI in DIS, which is different form the traditional interactive UI. Fundamentally, the paper focuses on improving the intelligence of query processing in UI according to introducing the method and technology of Artificial Intelligence and implements a friendly and intelligent in part processing module of UI. Finally, the specific details of the paper are presented as follows:1. Proposing an ontology-based UI method of data integration. Taking ontology as the uniform data view. Using OWL language to describe data source and construct the model of domain knowledge. Solving the deficiency of semantics in query process according to semantic framework of ontology.2. Designing and implementing an intelligent user query interface. The interface combines so many technologies such as ontology, Description Logic(DL), Chinese Segment, OWL-S to deal with users'query. Especially, the original querys, which were keyword-based and ill-formed, will have been expanded into a new query set which include the associated concept in semantic framework of ontology. Finally, the intelligent user query interface will export normative query request cell. Presently, the interface module can achieve certain function such as segment, semantic processing and output of standard query cell.3. Participating in development of the visual editorial tool of ontology. This tool and DL concept classification algorithm will be fused in the interface module of DIS.