Object Centric Process Mining

Object-centric process mining extends traditional process mining by explicitly modeling the interactions of multiple objects within a process, moving beyond the single-case assumption. Current research focuses on developing novel graph-based representations, such as heterogeneous object event graphs, and applying graph neural networks for tasks like predictive monitoring, anomaly detection, and conformance checking. This approach offers more accurate and nuanced process understanding, leading to improved process optimization and enhanced decision-making in diverse domains, particularly those involving complex, multi-object interactions.