Hsin-Hsi Chen and Cheng-Hsih Chang
Department of Computer Science and Information Engineering
National Taiwan University
Taipei, Taiwan 10764, R.O.C.
Time is indispensable for many computer applications because of their highly frequent use in the real world. How to represent and manipulate temporal knowledge adequately and efficaciously in computers has become a critical issue. Temporal logic is one of these useful formalisms due to its perspicuous, concise, and inferable properties. Time in temporal logic is usually classified into two different structures, i.e., time point and time interval. Several temporal logic languages based on different time structures have been presented. TEMPURA, TOKIO, and TEMPLOG are suitable only in time point structure. TIMELOGIC uses time interval, but it only considers the precedence relations of a set of actions. All of them are weak in inferring durations when properties remain true and in specifying the relations between durations. This paper will propose a practical tool IB-TEMPLOG for temporal knowledge management. It adopts interval structure. Knowledge in IB-TEMPLOG is represented by temporal facts and rules. The former specifies the occurrence of events or the durations when properties persist true. The latter defines the temporal relationships and constraints among events or properties. A proof procedure is given to derive new temporal knowledge from the basic one.
Keywords: interval calculus, knowledge representation, logic programming, non-monotonic reasoning, temporal logic
Received January 17, 1991; revised August 26, 1991.
Communicated by Ferng-Ching Lin.
*Parts of this paper were presented at the 4th Australian Joint Conference on Artificial Intelligence, Perth, Western Australia, November 1990.