An interactive approach for group-based event log exploration

A major goal in process mining is to analyze processes to determine possible improvements. However, event logs often bear substantial complexity, posing challenges for process analysts. Consequently, analysts often split event logs into more serviceable groups. While tool support is a crucial enabler for this task, and many approaches for event log analysis are available, a gap remains regarding tools and methods for organizing and structuring event logs. To address this gap, we propose the Case Group Explorer, an approach to support event log grouping using interaction and visualization computationally. We instantiate our artifact as a software prototype and evaluate it through a competing artifact analysis, the application on several event logs, and a user study involving 13 practitioners. Thus, we contribute by creating design knowledge for event log exploration and process group analysis at the intersection of process analysis and visual analytics. • In addition to prevalent process mining solutions that mainly focus on control flow, analysts aim at identifying relevant groups for initial analysis. • An approach is proposed to support interactive event log analysis by forming groups of process traces with aggregated features. • The feasibility, applicability, and ease of use of this approach are demonstrated.