Metadatabase Meets Distributed AI

Heterogeneous Distributed Database Management Systems (HDDBMS) involve the interoperability of data sources. One approach to achieve this type of integration is to build interfaces between the different databases being integrated. This approach holds, for a particular case, at a specific point in time. In this case however, the database structures need to be adapted. Such adaptation is not advisable since the local systems are usually important for their organizations. Therefore, an integration model that assures flexibility and scalability must be based on some knowledge of the underlying model of the different local databases. One solution is the use of the metadata concept, as a means to describe the logical and physical data characteristics. The metadata concept leads to the development of a Metadatabase system, which is viewed as a knowledge base about the local systems. The Metadatabase work at Rensselaer Polytechnic Institute (Troy, New- York) [11] and Universite Laval (Ste-Foy, Quebec) [2] has focused on creating such an integration environment and on defining its principal components. These solutions have been developed outside the context of Distributed Artificial Intelligence (DAI) and would certainly benefit from the results in that field of research. In this paper, we explain how the Metadatabase approach can be mapped into or associated with DAI concepts, and how it could benefit from techniques and theories pertaining to the DAI field.