Tim Kraska Receives A Sloan Research Fellowship


For more content about Tim Kraska, click here.

Picture yourself in a meeting with colleagues, looking up at a conference room wall. Not far in the future, Assistant Professor Tim Kraska of Brown University's Department of Computer Science (Brown CS) expects it to be equipped with an interactive whiteboard that will enable domain experts and data scientists to work together in a single meeting to visualize, transform, and analyze complex data in real time. Among many other benefits, the process would remove the need for multiple days of back-and-forth interactions. Between our present and this future is a complete rethinking of the full analytics stack, from its user interface to its smallest components, and incorporating pertinent algorithms. 

Tim has just been named an Alfred P. Sloan Research Fellow in one of the oldest and most competitive fellowship programs in the country. He's the eighth faculty member to receive the honor, which Brown CS has now received for four years in a row. The fellowships, which take the form of a $50,000 grant used over a two-year period, honor and promote the science of outstanding researchers early in their academic careers who show outstanding promise for fundamental contributions to new knowledge.

"Very few people," Tim says, "possess a strong domain expertise and a deep understanding of machine learning, data management, visualization, and many other related fields. My research aims to democratize data science by enabling a broader range of users to unfold the potential of their data through the development of a new generation of algorithms and systems for interactive and sustainable data-driven discovery."

Tim's research that the Sloan fellowship will help fund has three major components: 

  • Tupleware, a parallel high-performance UDF processing system designed for “normal” users, not the world's Googles and Microsofts. By avoiding unnecessary coordination overhead and new optimization techniques for modern hardware and smaller clusters, Tupleware can achieve up to three orders-of-magnitude performance improvement over alternative systems for common analytical tasks.
  • Vizdom, a touch-and-pen-based interface built on top of Tupleware, providing users with a visual interface in which they can quickly test out a wide variety of hypotheses with minimal effort as they interactively refine these hypotheses over time. Working with the Brown CS Graphics group, Tim's researchers are developing new data structures and algorithms especially tailored towards the needs of visual data exploration and integrating them into Tupleware in the form of an Interactive Data Exploration Accelerator (IDEA).
  • Sustainable insights: While visual data exploration tools are key to democratizing data science, they also introduce new risks. It's easy to mistake a visualization for a statistically significant fact, even though it might just be a random occurrence. Together with Professor Eli Upfal of Brown CS, an expert in statistical methods, Tim is working to develop new techniques to control the multi-hypothesis error in the context of visual data exploration. 

For more information, click the link that follows to contact Brown CS Communication Outreach Specialist Jesse C. Polhemus.