ABSTRACT
In this paper, we focus on the representation of, and the reasoning with, metric bounds with possibilistic distributions. Metric bounds have been used to represent the prepositional quantitative relations between events. For example, the ACM student chapter meeting will take at least 45 minutes and at most 65 minutes is a metric bound on the event meeting. Metric bounds use two extreme bounds for duration of an event even though it rarely occurs. In this paper we show how to associate possibilistic distribution with the metric bounds, and thus makes the model more accurate. When we associate the possibilistic distribution with metric bounds, the composition of metric intervals becomes quite complex as we have shown in this paper. For computational convenience, we provide a method to approximate the distribution and show how to compute the composition of complex composite distributions. The representation can be further enriched by incorporating temporal transition rules. We discuss about the representation and the propagation of temporal constraints using the rule. We illustrate our representation and reasoning mechanism with an example.
