Dynamic Path Consistency for Interval-based Temporal Reasoning

Path consistency is an important component of the resolution method needed to check for the consistency of an Allen’s Interval-based temporal network. While this local consistency algorithm reduces, in general, the size of the search space before applying the backtrack search, it implies the consistency of the problem for some specific temporal networks. Our goal in this paper is to maintain the path consistency of a temporal network in a dynamic environment i.e anytime a temporal relation is added or removed. For this purpose we propose a dynamic path consistency algorithm based on van Beek’s path consistency technique. Experimental tests on randomly generated dynamic interval-based temporal networks demonstrate the efficiency of our algorithm to deal with large size temporal problems in a dynamic environement.