In this paper, we develop a historical relational data model by extending the standard relational data model with operations on temporal data. The historical relational data model uses the time point as the basis for representing both events and states. We store an event occurrence as an interval bounded by lower and upper time points, and we store a state as an interval bounded by start and stop events. The data model thus provides interval stamping of tuples in the relations for both events and states. We develop a set of temporal operations on time points and intervals as the basis for understanding a proper relational algebra for historical relations. We also show how the operators in the historical relational algebra are used to execute an extended SQL for the historical relational database.
