Data Visualization at Khoury College of Computer Sciences
Building the tools that help us see and understand the world
How do we make sense of the onslaught of data we face every day? From tracking a daily run, to peering inside a virus, we increasingly depend on models and animations. The study of data visualization investigates how to connect the most complicated capacity of humans — our ability to take in visual information — with the streams of data that now define our world. How can we identify patterns, find gaps, understand trends — all within a fraction of a second? And how do we create images and animations that we increasingly depend on for work, daily life, health, and safety?
Pushing frontiers and shaping the future
It’s a challenge that Khoury College is tackling head-on through an interdisciplinary team of faculty — the largest such group in the US — bringing expertise from computer science, psychology, art and design, human-computer interaction and other fields. Together, these researchers push the frontier on how humans interact with and interpret data from computers—doing the research that will shape the future of visualization.
Sample research areas
- Perception and cognition
- Data storytelling
- Analytic provenance
- Exploratory data analysis
- Coordinated views and interactions
- Tree/hierarchical data
- Network data
- Multidimensional data
- Geospatial data
- Uncertain, missing, erroneous data
- Temporal event sequences
- User interface design
- Interaction design
- Evaluation methodologies
Domains of interest
- Health care diagnostic and treatment decision support, including diabetes, neurology, ophthalmology
- Cybersecurity, including unmanned autonomous system analysis
- Astronomy and physics, including 3D visualization and multidimensional data
- Computer science, including work in databases, programming languages, and systems
- Digital humanities, including networks of concepts in humanities texts, text and timeline visualizations
- Epidemiology, including the spread of infectious disease
Current project highlights
Dashboard design that supports data collaboration
Khoury Data Visualization faculty are working on ways to make digital dashboards, now part of many everyday tools more useful. They are researching a new approach to design dashboards, one that supports a collaborative conversation between user and data, instead of focusing only on one-way visual presentation. The guidelines they are developing and testing could help developers build dashboards that would support richer interaction with and analysis of data.
Blending data visualization and art to tell environmental stories
Eco-Mending is a graduate student research project at Khoury that combines historical imagery with contemporary data to tell stories of environmental damage and recovery. With case studies including oil pollution in Nigeria and ozone depletion, the project combines computer data visualizations, photography and physical constructions that capture past history and help tell the story of ‘eco-mending,’ stewardship to aid environmental recovery.
Deepening understanding of visualization literacy and addressing gaps
Data visualizations are everywhere now, but people range in their ability to use these visualizations to understand data and identify patterns: just as literacy determines how much someone can get out of a text, visualization literacy measures how much one can get from a data presentation. It’s important to understand how to measure this, as it can help guide the creation and presentation of essential data. Khoury researchers have found that current methods to measure visualization literacy are inadequate, and have developed a more nuanced approach that taps learnings from other literacy research, and reflects educational level and task complexity.
Recent research publications
Barriers to Photosensitive Accessibility in Virtual Reality
Authors: Laura South, Amy Pavel, Caglar Yildirim, Michelle A. Borkin
Virtual reality (VR) is becoming prevalent, moving beyond its current niche use in gaming and simulations. But VR data visualizations present hurdles for people with photosensitivity who may not be able to use it safely. Khoury research identified the specific challenges that make VR less accessible for this group, enabling software developers to create a safer and more inclusive experience.
Struggles and Strategies in Understanding Information Visualizations
Authors: Maryam Rezaie, Melanie Tory, Sheelagh Carpendale
Khoury researchers are exploring the challenges people face when interpreting infographics. Through observation of actual users, the researchers identified specific design flaws in color use and labeling, and other factors that cause confusion. This research helps designers create visualizations to enable users to make sense of visualizations and understand the underlying data accurately.
State of the Art of Visual Analytics for eXplainable Deep Learning
Authors: B. La Rosa, G. Blasilli, R. Bourqui, D. Auber, G. Santucci, R. Capobianco, E. Bertini, R. Giot, M. Angelini
Deep learning (DL), a type of machine learning focusing on neural networks with multiple layers, is at the heart of advances in AI, but unfortunately is very difficult to understand – even by the developers who create the underlying software code. Typically DL works as a ‘black box’ in which input and outputs are visible, but the actual processes are unknown. Researchers at Khoury College are using Visual Analytics to reveal what’s going on with these processes and provide ways to explore and understand them.
-
Khoury VIS Lab: Jane Adams, Uzma Haque Syeda, Carey Barry, Michael Correll, Racquel Fygenson, Connor Wilson, Shani Spivak (back row); Mahsan Nourani, Cody Dunne, Lace Padilla, Laura South, Sydney Purdue, and Mac Creamer (front row). -
Eduardo Puerta (left) and Connor Wilson (right) discuss 3D printing and network layouts with their advisor Cody Dunne. -
Michelle Borkin demonstrates an augmented reality data visualization application for space imagery. -
Lace Padilla and PhD student Racquel Fygenson discuss hurricane forecast visualizations. -
Cody Dunne presents an advanced network visualization. -
3D model of a sign for a study about 3D printing braille. -
PhD student Jane Adams illustrates differential co-expression of genes in mice under control vs. experimental treatment as part of her doctoral research collaboration with the Jackson Laboratory's Center for Systems Neurogenetics of Addiction.
Related labs and groups
Faculty members
-
Lace Padilla
Lace Padilla is an assistant professor at Khoury College, jointly appointed with the College of Science. She combines cognitive, neural, and computer sciences to study how people process data visualizations and how those visualizations can communicate more accurately and effectively.
-
Dakuo Wang
Dakuo Wang is an associate professor at Khoury College. He is also an ACM Distinguished Speaker and gives talks around the world on his research into human-centered AI (HCAI) systems.