Clustering

In this case study, we show that visuEL allows visualizing clusters of traces quickly. We took the dataset "BPI 2020, Permit Log" and created 12 clusters using Kmeans. The original dataset is visible in blue. Using the visualizations, we claim that it is possible to extract some quick insights. For instance, some clusters are more complex than others in terms of behaviours. We believe that it should help process analysts to refine the clustering algorithms and/or leverage specific process mining techniques for some clusters.

Legend