- Computational Social Science
- Critical Data Studies
- Data Science
- Economics and Information
- Education Technology
- Ethics, Law and Policy
- Human-Computer Interaction
- Human-Robot Interaction
- Incentives and Computation
- Infrastructure Studies
- Interface Design and Ubiquitous Computing
- Natural Language Processing
- Network Science
- Social Computing and Computer-supported Cooperative Work
- Technology and Equity
Please join us for the Information Science Colloquium with guest, "Leo" Zhicheng Liu, a postdoctoral scholar in the Department of Computer Science at Stanford University. He obtained his Ph.D. in Human-Centered Computing from Georgia Institute of Technology in 2012. His research focuses on developing tools to enable data enthusiasts to more effectively perform visual data analysis: designing novel interfaces for data modeling, building user-centered visual analytic systems for domain experts, and developing computational methods for scaling interactive visualizations to big data. He is a Foley Scholar and a recipient of the Doctoral Dissertation Award from the College of Computing, Georgia Tech.
Title: Enabling Data Enthusiasts: Visual Analysis Tools for Big Data
Abstract: Interactive visualization can be an effective tool for both non-technical enthusiasts and data scientists to understand, analyze and present data. Current visualization tools, however, are often insufficient. Instead of focusing on their analysis questions, users spend time munging data, figuring out how to specify visualizations (sometimes with ineffective results), or struggling to cope with large data volumes.
In this talk, I discuss research on novel interactive visualization systems that address these challenges. To aid articulation of intended visualization semantics, our Ploceus system explores the problem space of transforming tabular data into networks and supports dynamic visual analysis of these networks through a direct manipulation interface. To achieve real-time visual querying of multi-dimensional data, we investigate the design space of scalable visual summaries and develop computational methods that reduce interactive latency. The resulting system imMens is among the first to scale visualizations for billions of rows of data while maintaining interactivity. Collectively, these approaches contribute to a larger vision of enabling data enthusiasts to effectively turning data into human-understandable knowledge and actionable insights.
A reception will be held immediately after.