Issue No. 03 - June (1995 vol. 10)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/64.393144
<p>We describe a four-level blackboard architecture that supports schema integration and provide a detailed description of the communication among human and computational agents that this system allows.</p> <p>Today's corporate information system environments are heterogeneous, consisting of multiple and independently managed databases. Many applications that assist decision making call for access to data from multiple heterogeneous databases. To facilitate this, there needs to be an integrated representation of the underlying databases that allows users to query multiple databases simultaneously. The process of deriving this integrated representation is called schema integration.</p> <p>Schema integration is time consuming and complex, as it requires a thorough understanding of the underlying database semantics. Since no data model can capture the entire real world semantics of each database's objects, this process requires human agent assistance. Although certain aspects of schema integration can be automated, interaction with designers and users is still necessary. In this article, we describe how blackboard architectures can facilitate the communication among human and computational agents for schema integration.</p>
V. Ramesh and S. Ram, "A Blackboard-Based Cooperative System for Schema Integration," in IEEE Intelligent Systems, vol. 10, no. , pp. 56-62, 1995.