Abstract: Causality helps people reason about and understand complex systems, particularly through what-if analyses that explore how interventions might alter outcomes. Although existing methods embrace causal reasoning using interventions and counterfactual analysis, they primarily focus on effects at the population level. These approaches often fall short in systems characterized by significant heterogeneity, where the impact of an intervention can vary widely across subgroups. To address this challenge, we present XplainAct, a visual analytics framework that supports personalized causal analysis by enabling interventions at the individual level within subpopulations. We demonstrate the effectiveness of XplainAct through two case studies: investigating opioid-related deaths in epidemiology and analyzing voting inclinations in the presidential election.
Teaser: The image shows the XplainAct interface:
We use an opioid dataset for illustration. (A) Choropleth map highlighting Coconino County, Arizona (bold black outline) and counties with similar opioid-related socioeconomic indicators with their respective opioid death rates colored according to the legend on the right; (B) Feature explanation panel using LIME to reveal the contribution of each socioeconomic factor to the opioid death rate in Coconino County; (C) Parallel coordinates plot comparing Coconino County’s multidimensional profile (red line) with its socioeconomic peers (blue lines) and highlighted on the choropleth map (A); and (D) Slider group for defining contextual similarity criteria to Coconino County’s profile.
Video: Watch it to get a quick overview how a crowd participant would build a small causal network with our interface:
Paper: Y. Zhang. K. Hegde, K. Mueller, “XplainAct: Visualization for Personalized Intervention Insights,” IEEE Visualization & Visual Analytics (VIS), Vienna, Austria, November 2025. PDF