Research On Ontology And SSH-Based Heterogeneous Data Integration

With the speed development of our economy and science research capacity, high-tech information technologies such as Cloud Computing, Big Data and Internet of Thing have been used in our daily life and work. One sharp contrast to this fast development is the severity of information island, which caused by the failure of information sharing, due to that some date source have different date naming and expressions, or the difference between date sources’ structures and modes.In order to solve above questions, researchers from aboard and home have conducted deep research of heterogeneous data integration, and put forward points such as Data Warehouse, Federal Database, Ontology-based Virtual Integration and CORBA-based Data Dictionary. As in recent years the focus of heterogeneous data integration has been shifted to semantic heterogeneity, Ontology-based heterogeneous integration became the research hotspot. This study also based on ontology and combined the system building idea based on SSH. This method makes it possible to expand and change the underlying data source, which is a big improvement to the system robustness compared to traditional methods, unifies the management of system components and enable the automatic extraction of data pattern of relational data source. Additionally, the components of web layer and data persistence layer are integrated by spring which reduces the coupling coefficient of system and program, and thus improves integration efficiency.The emphasis of the research is to construct the local ontology which can map ea ch data schema of heterogeneous data source and the global ontology which consists o f local ontology and domain ontology, as well as the global view which takes the glob al ontology as the integrated system. In this case, domain ontology which is regar ded as the constructing prototype of global ontology is constructed by protege. Hibern ate reverse engineering is used to extract the relation schema of heterogeneous databa se automatically, and then generating OWL ontology files, after which, Jena develop ment kit will be used to extract the eigenvalues of local ontology and domain ontology. Ontology computation algorithm has been promoted in this research, thus changing t he computation algorithm that only relies on mature knowledge system. At last, tests are taken to prove the rationality and advantage of the research.