Temporal Data Exchange

Abstract

In this work, we study data exchange for temporal data. There are two views associated with temporal data: the concrete temporal view, which depicts how temporal data is compactly represented and on which implementations are based, and the abstract temporal view, which defines the semantics of temporal data. Based on the chase procedure, which is a fundamental tool in relational data exchange, two new kinds of chase are proposed in this paper: the abstract chase for the abstract temporal view and the concrete chase for the concrete temporal view. While labeled nulls are sufficient for relational data exchange, they have to be refined in temporal data exchange to keep the connection between the result produced by the concrete chase and the result of the abstract chase. We show that the concrete chase respects the semantics defined by the abstract chase and provides a foundation for query answering