Research And Implementation Of Ontology-based Heterogeneous Data Integration System

In recent years, the amount of data and information within many large enterprises and institutions has increased continuously, and these data tend to be more and more heterogeneous and distributed, etc. With the accelerated pace of informatization construction, demands on integrating and sharing these heterogeneous data has become more and more urgent. This paper takes heterogeneous data sources as the object of the study, builds a heterogeneous data integration system architecture model, studies the construction methods of a global ontology , improved local ontology construction method as well as global query decomposition algorithm and develops the ontology-based heterogeneous data integration prototype system aiming at academic building needs of a university. The system integrates three heterogeneous data sources involving the academic building and provides an effective way for users to access data resources in a uniform way. Main research contents and conclusions are as follows:(1) This paper takes the construction of the global ontology and the local ontology as the goal and studies the ontology construction method, improved rules based on relational database extraction ontology and based on XML files. Data sources involved in this paper are the structured relational database MySQL and Oracle as well as the semi-structured data files XML, and local ontology semantic models are respectively constructed in this paper aims at different forms of the three data sources by adopting different construction rules. Besides, the global ontology semantic model is constructed with the participation of experts in the field in allusion to global application. Finally, the Protégéontology modeling tools are adopted to build the global ontology and the local ontology.(2) Based on the construction of the global ontology and the local ontology, this paper adopts manual methods of building ontology mapping rules to build mapping information from the global ontology to the local ontology as well as the mapping information from the local ontology to the data source and provides support for query decomposition and query conversion. In the construction of the mapping information from the global ontology to the local ontology, the class, the data type association and the individual association of the global ontology need to be mapped to the class, the data type association and the individual association which correspond to each local ontology; and in the construction of the mapping information from the local ontology to the data source, the class and the data type association as well as the individual association of the local ontology need to be mapped to tables, attributes and foreign keys of the relational database.(3) The paper studies the problem of the query conversion and improves the query decomposition algorithm and the query conversion algorithm. In the paper, the query decomposition algorithm is used to decompose SPARQL statements which aim at the global query into SPARQL statements which aim at the local ontology, and then the query conversion algorithm is used to convert the SPARQL statements which aim at the local ontology to SQL statements which aim at the data source.(4) Taking the heterogeneous data in the field of the academic information of a university as the background, the paper develops the ontology-based heterogeneous data integration prototype system, thereby realizing effective integration of the heterogeneous data and providing the users with a query interface which enables the users to simultaneously access the information of the respective data sources in the uniform manner.