LAUSR

Time modeling is an important issue in information management. Both traditional and soft computing approaches have been thoroughly studied in the existing literature. Traditional approaches are specifically designed to efficiently deal with perfect temporal data. Soft computing techniques additionally support the efficient handling of imperfect temporal data. In this paper, a novel soft computing technique is presented, based on fuzzy set theory and using possibility theory, to represent, handle, and reason with time intervals that have uncertain starting and/or ending boundaries. Hereby, a 2-D visualization approach, which is called the triangular model, is generalized to efficiently and effectively cope with such time intervals subject to uncertainty. The proposed approach provides a straightforward and compact visualization for possibilistic interval data, which supports temporal analytical processing and in which temporal distributions are easy to observe, explore, and analyze. Moreover, it is shown how temporal analysis can be carried out using the proposed novel approach. In this context, both the presence and absence of uncertainty in one or more time intervals are supported. The value of the proposed novel approach is demonstrated through the development of two illustrative use cases.