Probabilistic Conformance Checking Based on Declarative Process Models

Abstract

Monitoring and conformance checking are fundamental tasks to detect deviations between the actual and the expected courses of execution in a business process. In a variety of application domains, the model capturing the expected behaviors may be intrinsically uncertain, e.g., to distinguish between standard courses of execution and exceptional but still conforming ones. Surprisingly, only very few approaches consider uncertainty as a first-class citizen in this spectrum. In this paper, we tackle this timely and challenging problem considering the setting where the model is described as a set of declarative, temporal business constraints. First, we delve into the conceptual meaning of probabilistic business constraints, and argue that they can be discovered from event data using already existing techniques. Second, we study how to monitor probabilistic constraints, where constraints and their combinations may be in multiple monitoring states at the same time, though with different associated probabilities. Third, we operationalize this monitoring framework using automata coupled with probabilities, and report on its actual implementation.

Stochastic Process Discovery by Weight Estimation

“…This contrasts with control-flow based process models, such as Petri nets used in our framework, which describe permitted behaviour. Techniques for automatic process discovery of probabilistic declarative models also exist [23]. Transforming the significant differences between the forms of control-flow and declarative models, and evaluating the result for stochastic conformance, put rigorous comparison beyond the scope of this paper.…”