Interval-valued data structures and their application to e-learning

Abstract

The paper is devoted to the problem of replacing crisp numbers with interval numbers in soft computations. The original concept of an interval-valued vector (IVV) is introduced, and the new extensions of classic similarity measures are proposed to handle IVV matching. Finally, the presented data structure and the matching methods are used in the process of an automated evaluation of tests in e-learning (distance learning within the Internet). © Springer-Verlag Berlin Heidelberg 2005.