Step-by-Step Case ID Identification based on Activity Connection for Cross-Organizational Process Mining

Cross-organizational process mining aims to discover an entire model across multiple organizations where their identifier (ID) systems are not managed uniformly, and each organization has independent ID system. been gaining popularity as information increase in complexity. However, previous methods have limitations that they do work well for event logs contain only common items, or cyclic orchestrations, which indicates the contains loops. In this paper, we propose accurate cross-organizational technique based on a step-by-step case identification mechanism uses items can handle orchestrations. Step-by-step repeats following steps: 1) of IDs activity connection adjacent pairs, 2) extraction additional connections by leveraging newly identified IDs. We alternately identify most probable pairs remove events belonging these from log, contributes extracting narrowing down candidates pairs. Evaluation using real-world showed proposed method generates with more than 98.4% precision 94.2% recall two datasets, outperforming methods.