Tuning Alignment Computation: An Experimental Evaluation

Conformance checking aims at assessing whether a process model and event data, recorded in an event log, conform to each other. In recent years, alignments have proven extremely useful for calculating conformance statistics. Computing optimal alignments is equivalent to solving a shortest path problem on the state space of the synchronous product net of a given Petri net and event data. State-of-the-art alignmentbased conformance checking implementations exploit the A∗-algorithm, a heuristic search method for shortest path problems, and include a wide range of parameters that likely influence their performance. In this paper, we present an exploratory empirical evaluation of parametrization of the A∗-algorithm used in alignment computation. Our initial results show that the performance of alignment computation greatly depends on adequate parametrization of the underlying search algorithm.