Declarative Process Mining with MINERful, Reloaded

Abstract

Declarative process specifications are defined by a set of behavioural constraints exerted over the execution of activities. These constraints are rooted in Linear Temporal Logic over finite traces. Declarative process mining encompasses a collection of techniques based on recorded process executions (event logs), with the objective of extracting such specifications and enhancing them. This paper presents MINERful, an open-source tool for declarative process mining in its new, reloaded version. It offers a range of functionalities revolving around the mining cycle of a process specification: its automated discovery from event logs, its simplification to remove redundancies and inconsistencies, its simulation to generate synthetic datasets, and fitness checking to gauge its level of conformance with logs. We showcase its usage with a real-world event log in the healthcare domain.