Data visualization plays a central role in process mining, particularly when analysts need to compare multiple process models, often representing different process variants, to assess changes, detect deviations, or benchmark performance. However, effectively supporting such comparative analysis remains challenging. To investigate this problem, we combine three perspectives: interviews with nine process mining analysts, established visual analytics principles, and an analysis of existing process mining visualizations. Based on this analysis, we identify key challenges for comparative analysis, including weak support for goaldriven analysis, insufficient scalability, and limited flexibility. Based on these insights, we propose an interactive approach for the flexible, scalable, and goal-driven comparison of multiple process models. The approach enables dynamic, goal-driven highlighting of differences and similarities, allowing analysts to tailor comparison based on specific analysis objectives. We developed a prototype implementing the main concepts and interaction mechanisms, and evaluated its perceived usefulness and usability through two workshops with a total of ten participants (students and experts). The results provide initial design insights for visual analytics techniques supporting process variant comparison and highlight directions for future research.

Top