Out of Babylon: Object-Centric Conformance Checking on Graph-Based Abstractions

Process mining employs data-driven techniques to analyse and optimise business processes. An important step in such an analysis is to identify and quantify deviations between the observed and modelled behaviour of a business process using conformance checking techniques. While traditional conformance checking approaches utilise process models that primarily focus on events, recently proposed approaches also consider the involvement of business objects. These object-centric process models introduce new behavioural deviations due to the involvement of objects in events like missing objects in the execution of an event. However, existing object-centric conformance checking techniques do not precisely detect all these deviations and can only be applied to specific object-centric modelling formalisms. In this paper, we introduce a novel formalism-agnostic object-centric conformance checking approach that presents a common denominator for many object-centric modelling formalisms, enabling comparable conformance checking results. We evaluate our approach on a public object-centric log by demonstrating its ability to detect deviations across multiple modelling formalisms.