Evidential Reasoning for Heuristic Algorithms to Solve Interval Algebra Network

Evidential Reasoning provides quantitative enhancements to the reasoning formalisms which use set based algebra. Interval Algebra is a Temporal Reasoning formalism which uses sets as representation forms and employs tools based on set operations. The traditional reasoning in the Interval Algebra is to apply a search process, Backtrack search, which takes exponential runtime. We augment Interval Algebra with Evidential techniques to make it more expressive and make use of the quantitative enhancements as heuristics to develop backtrack free algorithms which are sound with polynomial runtime. We evaluate efficiency of the algorithms experimentally by generating random instances of the Interval Algebra network. The percentage of success is greater than 90 for the best of these algorithms and its runtime is very low when compared with that of backtrack search method.