| At present, both between the various enterprises in the same industry and between same enterprises with different professional, there are a variety of heterogeneous data sources. They are isolated from each other, it’s difficult to realize the data exchange and sharing between them. Enterprises generally have ‘information island’ problems that difficult to solve.Therefore, people have proposed many solutions for data integration. However, because there is no standard to follow the plan, in other words, there is no unified and shared canonical‘metamodel’ to solve the problem of data model semantic problem and mapping between heterogeneous data models semantic problem. So these plans generally lack of unity,reusability and standardization.At many traditional methods for research on information integration, there are problems of weak semantic interaction. To solve these problems, the International Organization for Standardization proposed a series of MFI(Metamodel Framework for Interoperability)standards. These standards gives the standardized ‘metadata model’ to describe these meta information in information model registration and the mapping information registration from source information model to target information model. These standard to solve the data interoperability problem fundamentally, especially,the semantic interoperability is enhanced.However, because the MFI standard is just a standard system framework, especially in the practical application, it is found that the mapping degree of MFI mapping registration standard can not meet the requirements of practical application, MFI only supports limited mapping type, and the complexity of the mapping is also very high. Therefore, this thesis extracts and defines complex mapping types through the in-depth study of MFI-10 and MFI-12 standards, and these complex mapping is extended to the mapping meta model with the practical application in Oilfield. Not only that, in order to make the user transparent and unified access to data in different data sources, this thesis introduce the MDR standard of ISO/IEC 11179 as a global public data model, and propose three tier technology framework system-unified registration model framework for data interoperability. Because the proposed model conforms to the international standard, it provides a scientific and normative model system for enterprise information integration.Even the model can be used as a integration standard for enterprise. In order to validate the three tier technology framework system, the interoperability between isolated data sources is realized by using the unified registration model framework for data interoperability with the practical application in Oilfield. Thisthesis is develop the query engine with reference to Mediator/Wrapper technology. This engine provides users with a query interface, so that users can query the data in the underlying model that is needed by the query instructions on the top level model to shield the underlying data source.In the field of petroleum, Daqing oil field has produced many kinds of heterogeneous data models such as E&P, EPDM and A5 in the long term of information construction. These models are independent of each other.Information exchange between these models has become a problem had plagued Daqing oil field for years. The registration model framework proposed in this thesis has been well applied in the information interoperability of data model in the oil field. The unified registration system is designed and developed in this thesis with the practical application in Oilfield. This system provides a registration interface for the registration administrator to manage the model information and mapping information between the models. At the same time, it also provides a query interface for the user to query the desired data. This system shows that the three layer model into these international standards can well solve the problem of enterprise data sharing. |
PDF Full Text Request