Efficient Evaluation of Similarity Quantified Expressions in the Temporal Domain

Abstract

Modelling of fuzzy temporal quantified statements is of great interest for real time systems. In [1] use of fuzzy proportional quantifiers to model temporal statements (sentences involving occurrence of events within a time framework) has been proposed. By using these proportional quantifiers a semantics can be associated to expressions like “medium or high temperature values were measured together to risky high pressure values in the last few seconds”. Nevertheless evaluation of a number of temporal statements cannot be modelled with these quantifiers, as “Association between risky high pressures and high temperatures has been very high in the last few seconds”. We will see how evaluation of these similarity or correlation expressions between two signals can be modelled by using similarity quantifiers.