Integration of different information sources, such as a set of databases (DBs), can be achieved by considering DBs as agents of a multiagent system. This is one of the basic intuitions of federated databases (FDBs). Differently from the case of a single DB, a completely satisfactory formal semantics for FDBs is still missing. The goal of this paper is to fill this gap by providing a semantics for FDBs which overcomes the drawbacks of the previous approaches. The semantics proposed in this paper, called Local Models Semantics for FDBs, is based on the intuition that the DBs of a federation are partial views on a common real world. We show by means of examples that Local Models Semantics copes with classical problems of FDB formalization, such as semantic heterogeineity, inter-schema dependencies, and query distribution
Local Models Semantics for Information Integration
Serafini, Luciano;Ghidini, Chiara
1997-01-01
Abstract
Integration of different information sources, such as a set of databases (DBs), can be achieved by considering DBs as agents of a multiagent system. This is one of the basic intuitions of federated databases (FDBs). Differently from the case of a single DB, a completely satisfactory formal semantics for FDBs is still missing. The goal of this paper is to fill this gap by providing a semantics for FDBs which overcomes the drawbacks of the previous approaches. The semantics proposed in this paper, called Local Models Semantics for FDBs, is based on the intuition that the DBs of a federation are partial views on a common real world. We show by means of examples that Local Models Semantics copes with classical problems of FDB formalization, such as semantic heterogeineity, inter-schema dependencies, and query distributionI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
