Precision Evaluation Methods of Petri Nets Modelling Discrete-Event Processes

Abstract

In process mining, conformance checking determines the quality of discrete-event process models through several measures computed with respect to a set of event sequences λ from which the model N is obtained. Precision is a measure that evaluates the surplus language of the model LN with respect to λ. This article outlines four precision evaluation approaches for Petri net (PN) workflow models obtained through process discovery techniques. First, the basics of the quality measures of process models stablished in conformance checking are summarised. Then, the methods for model precision evaluation in literature are outlined and analysed. Finally, the advantages and limitations of the methods are pointed out and compared.