Options
2020
Conference Paper
Titel
Conformance checking approximation using simulation
Abstract
Conformance checking techniques are used to compute to what degree a process model and real execution data correspond to each other. In recent years, alignments have proven to be useful for calculating conformance statistics. Most alignment techniques provide an exact conformance value. However, in many applications, it suffices to have an approximated alignment value. Specifically, for large event data and using standard hardware, current alignment techniques are time-consuming and sometimes intractable. This paper proposes to use simulated behaviors of process models to approximate the conformance checking value. To simulate a process model, we exploit the behavior in the given event data. This method is independent from the process model notation and provides upper and lower bounds for the approximated alignment value. We assess the quality of our approximations and compare it to existing approximation techniques. The experiments on real event data show that using the proposed method, it is possible to achieve significant performance improvements.