A Process-Oriented Approach to Analyze Analysts' Use of Visualizations: Revealing Insights into the What, When, and How
Journal
Computer graphics forum
ISSN
0167-7055
ISSN-Digital
1467-8659
Type
conference contribution
Date Issued
2025-05
Author(s)
Abstract
Despite Visual Analytics (VA) tools being essential for supporting data analysis, evaluating their use in real-world analytical processes remains challenging. Traditional evaluation methods often overlook the nuanced and evolving nature of analysis processes and are not always suitable for investigating scenarios in which analysts combine multiple tools and visualization types. In this paper, we propose a flexible analysis approach for studying analysts’ use of visualizations within and across VA tools. Our qualitative method allows researchers to extract user behavior and cognitive steps from screen recordings and think-aloud data and generate event sequences that capture analytic processes. This enables the analysis of usage patterns from multiple perspectives and levels of granularity and allows for the evaluation of effectiveness measures, such as efficiency and accuracy. We demonstrate our approach in the domain of process mining, where our findings provide insights into the use of existing visualizations, and we reflect on lessons learned from this application.
Funding(s)
This work was funded by the Swiss National Science Foundation as part of the ProMiSE project under Grant No.: 200021\_197032
The work was supported, in part, by the Swedish Research Council under Grant No.: 2020-05000
Language
English (United States)
Keywords
Empirical studies
Evaluation
Visualization design and evaluation methods
Process Mining
Visual Analytics
HSG Classification
contribution to scientific community
Refereed
Yes
Volume
44
Number
3
Event Title
EuroVis 2025
Subject(s)
Division(s)
File(s)![Thumbnail Image]()
![Thumbnail Image]()
Name
Supplementary Material.pdf
Size
974.32 KB
Format
Adobe PDF
Checksum (MD5)
2f3e0b8d308669e2714a676062d10887
open.access
Name
A process oriented approach to analyze analysts use of visualizations_preprint.pdf
Type
Main Article
Size
3.58 MB
Format
Adobe PDF
Checksum (MD5)
b04c27925290acf9243402a519ff5318