Project: Identifying and Reducing Visual Clutter for Process Graph Layouts
Description
Process mining tools help businesses gain insight into their processes by, among other things, providing automated visualizations of their process event logs. Process graphs play an important role in process analysis by providing an interactive graph layout in which nodes represent activities, and directed edges represent transitions between these activities. Generally, the layout of a process graph is constructed using a layered graph layout algorithm, created using the Sugiyama framework.
Graph readability, however, can be severely impacted by visual clutter. We survey process mining experts in order to find out, in the context of process mining, which visual aspects impact graph readability most. From these interviews several design decisions for edge routing are extracted, as well as a motivation for the visual bundling of edges. Related publications often focus either exclusively on edge routing, or do focus on edge bundling while failing to take the added visual clutter into account, which is often introduced as a result of overlapping edges.
We propose a novel, edge segment based, geometric edge routing scheme which minimizes edge crossings between consecutive layers. The edge routing algorithm reduces visual clutter by the application of design recommendations that originate from the interviews. The edge routing algorithm furthermore allows for segment-wise edge bundling using approximate offset Bézier curves, without introducing additional edge overlap.
After implementing our edge routing scheme into the UiPath Process Graph, process mining experts participated in an evaluative survey in which they were tasked with comparing process graph layouts generated using our edge routing scheme to process graphs layouts generated using the original edge routing implementation of UiPath. The evaluative survey shows that, statistically, process graph layouts generated using our edge routing scheme outperform the original process graph layouts.
Details
- Student
-
WBWilco Brouwer
- Supervisor
-
Huub van de Wetering
- External location
- UIPath
- Link
- Thesis