A probabilistic semantics for process mining

In declarative Process Mining (PM), accounting for uncertainty is essential to accurately model real-world business processes. Up to now, most traditional approaches have overlooked the possibility of integrating probability into process management. Starting from our previous works on this topic, we present an extension to our semantics that underlies a probabilistic declarative framework for PM, in such a way that we can manage uncertainty at multiple levels, from individual events to entire logs, by assigning probabilities reflecting a degree of belief or confidence in them. This framework is based on the Distribution Semantics of Probabilistic Logic Programming.