Representation of Time and Sequences of Events

Abstract : This paper describes a new approach to representing and reasoning with temporal information. A wide variety of temporal specifications can be converted into linear inequalities relating the endpoints of the events. Linear programming is then used to represent these constraints and perform deductions. The information is modularized into semantically related clusters of events each with its own tableau and related to each other by a reference frame transformation. This provides a uniform formally adequate representation which is complete and also computationally efficient. (Author)