Alexander Endert
Georgia Institute of Technology, GA·Computer Science
Software for exploring data increasingly includes artificial intelligence (AI) that suggests charts, highlights patterns and even nudges users toward conclusions. This help can be useful, but it is unclear when AI guidance improves people's decisions and when it leads them to over-rely on suggestions and miss important insights.
- Fields
Computer Science
Psychology
- Topics
visual analytics
data visualization
human-computer interaction
- Grants
- 1 known grant, through Sep 2029
What they work on
- Build an open-source software testbed with modular AI guidance components for chart creation and visual analysis of tabular data.
- Develop computational metrics that measure, in real time and after a task, how much a person relies on AI guidance, to detect both under-reliance and over-reliance.
- Run controlled user studies with the testbed to find which visual analytics tasks are most and least suited to AI guidance and to set baselines for appropriate use.
- Design and evaluate interface components that use reliance metrics to detect poor use of AI guidance and prompt users toward more effective engagement.
More about the research
This project studies how different kinds of AI guidance change the way people analyze data, so that future tools keep humans meaningfully in control.
Why it matters
People use visual data analysis software to make decisions in business, healthcare and national security. Knowing when AI suggestions help and when they cause over-reliance could guide designers in building tools that improve decisions while keeping humans in control. The open-source testbed could support other researchers and developers studying AI-assisted analysis.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Robert Krueger
New York University, NY·Computer Science·Early-career
Design scalable algorithms that jointly represent tissue images and omics data from spatial profiling, so that very large multimodal datasets can be explored interactively.
Jessica Hullman
Northwestern University, IL·Computer Science
Conduct formative studies with practicing analysts to characterize how they currently reason about alternative models and analysis steps, and where existing tools fail to support defining or comparing those alternatives.
Haeyong Chung
University of Alabama in Huntsville, AL·Computer Science·Early-career
Analyze existing multi-display applications in the academic literature to derive design knowledge on how multiple displays and devices can form one coherent visualization system for visual analysis.
The school
Georgia Institute of Technology: 50 labs on PI Pages, $1.5B a year on research, 572 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.