Improved Algorithms for Reasoning about Qualitative Temporal Constraint problems

Abstract

Representing and reasoning about incomplete and indefinite qualitative temporal information is and essential part of many artificial intelligence tasks. An Interval Algebra (IA) has been proposed by Allen as a model for representing qualitative temporal information about the relationships between pairs of intervals. In this paper, we present some algorithms to improve the tasks of reasoning with IA. First, we have improved Allen’s path consistency algorithm by eliminating duplicate computations and by applying tightest constraint first heuristic that leads to faster convergence. Second, for finding a consistent scenario, we show that a backtrack algorithm with backward checking and simple back jumping techniques, yields better performance than old one. By computational experiments, we test the performance of our algorithms on a set of randomly generated consistent networks. The results show that the overall average execution times are better in our method.