Are you eager to research and develop Visual Analytics techniques for Event Sequence exploration and analysis? Are you curious how to combine and explore the synergy between Visualization and Process Mining? In this project we aim to transform the way we analyze and visualize complex processes.
Information The field of process mining, a branch of business process management, revolves around discovering, analyzing, and optimizing business processes. It does so by extracting insights from event logs, structured records that capture sequences of activities, such as steps in a loan approval workflow, product purchases by customers, or patient journeys across hospitals. However, traditional process mining techniques focus primarily on control-flow, which describes the chain of activities. This singular focus often neglects other dimensions of processes, such as the time between activities and the time required for specific steps, the spatial distribution of tasks across locations, or the intricate relationships between resources and attributes involved in the processes. For example, in a supply chain scenario, understanding delays, warehouse locations, and inventory attributes is crucial, but often omitted in basic control-flow analysis. Process mining, as it stands today, is primarily based on computational techniques and algorithms to analyze and optimize processes. Methods such as process discovery, conformance checking, and performance analysis rely on algorithmic approaches to extract patterns, detect deviations, and identify inefficiencies from event logs. Although these techniques have proven effective in providing data-driven insights, they often fail to use visualization as a powerful means of interpretation and communication. Most process mining tools provide static and linear representations, such as process models (DFGs, BPMN) or Gantt charts, which are primarily suited for experts familiar with these abstractions. This limited use of visualization overlooks the potential for dynamic, interactive visual tools that can make complex process data accessible to a broader audience, enabling intuitive exploration and deeper understanding of multi-faceted process attributes, such as time, resources, and contextual dependencies and correlations.
In contrast, the field of visual analytics focuses on developing tools and techniques to support the interpretation of complex, multi-dimensional data. It provides concepts and techniques to explore and analyze event sequences, allowing patterns and anomalies to emerge intuitively. For example, visualizations such as timelines, Sankey diagrams, or network diagrams make it possible to uncover dependencies and bottlenecks in event sequences. Despite their shared goal of deriving actionable insights from event sequence data, process mining and visual analytics have largely operated in parallel, leaving significant opportunities for synergy unexplored.
By integrating the two domains, process mining can benefit from visual analytics’ expertise in handling multi-faceted data. Visualizations that incorporate temporal, spatial, and relational aspects can unlock deeper insights into processes. Conversely, visual analytics can leverage process mining techniques to enhance guidance and evaluation during data exploration. For instance, capturing event logs while users interact with visualization tools can provide valuable feedback to refine analytical techniques.
The challenge lies in merging these disciplines effectively, combining their strengths to fully realize the potential of multi-faceted data for visual process analytics. This project tackles this open problem, aiming to create a unified framework that harmonizes these complementary fields, transforming the way we analyze and visualize complex processes.
In this project, the focus is on developing Visual Analytics techniques for event sequence exploration and analysis to enable Multi-Faceted Visual Process Analytics. The project is performed within the Visualization cluster under the supervision of dr.ir. Stef van den Elzen and dr. Fernando Paulovich.
The visualization cluster (
https://research.tue.nl/en/organisations/visualization-3) at TU/e has a strong track record in visualization and visual analytics for high-dimensional data. It has generated several award-winning contributions at major visualization conferences (IEEE VIS, IEEE InfoVis, IEEE VAST, EuroVis); several successful start-up companies (MagnaView – now UiPath and SynerScope); and a number of techniques that are used on a large scale worldwide.