Process Mining Algorithms for Clinical Workflow Analysis

Process Mining focuses on the extraction of knowledge from data generated and stored by information systems. Log level data contains the signatures of executed processes. Many case studies have used process mining to discover the processes for compliance or identifying anomalies in business workflow. In this paper, we apply workflow analysis for a possible clinical setting by leveraging an open source data mining tool named ProM. We apply four available mining algorithms (Alpha, Heuristic, Inductive and Fuzzy miners) and evaluate the outputs that describe a workflow. The work provides a genesis for clinical practitioners on the advantages and disadvantages of applying various algorithms towards a dataset based on event logs.