log in  |  register  |  feedback?  |  help  |  web accessibility
Logo
ProgressiVis: a New Workflow Model for Scalability in Information Visualization
2105 Hornbake Building, South Wing
Thursday, September 10, 2015, 12:00-1:00 pm Calendar
  • You are subscribed to this talk through .
  • You are watching this talk through .
  • You are subscribed to this talk. (unsubscribe, watch)
  • You are watching this talk. (unwatch, subscribe)
  • You are not subscribed to this talk. (watch, subscribe)
Abstract

Information Visualization (infovis) has, for years, been limited to small data: a typical infovis application will work well with up-to 1000 items/records, a few can scale to 100,000 items, and very few, including the leading commercial products such as Tableau and Spotfire, have been able to deal with millions of items. Billions are seldom mentioned in the infovis literature. In contrast, the research fields of machine learning and databases are routinely dealing with datasets of several billions of items, and the numbers are growing.There are legitimate reasons why it takes time for infovis to start catching-up with these large numbers, and some work such as Lins et al. Nanocubes (http://www.nanocubes.net/) and Liu et al. imMens (http://idl.cs.washington.edu/papers/immens), have started to show possible routes to scalability. However, they both rely on either pre-computed aggregations that need hours to compute for large datasets, or on a highly parallel infrastructure performing aggregations on the fly. In my talk, I will explain why we need more flexible solutions and present a new workflow architecture called ProgressiVis, to achieve progressive computations and visualization over massive datasets.

Bio

Jean-Daniel Fekete is Senior Research Scientist (DR1) at INRIA, the French National Research Institute in Computer Science. He received his PhD in Computer Science in 1996 from Université Paris-Sud. From 1997 to 2001, he joined the Graphic Design group at the Ecole des Mines de Nantes that he led from 2000 to 2001. He was then invited to join the Human-Computer Interaction Laboratory at the University of Maryland in the USA for one year. He was recruited by INRIA in 2002 as a confirmed researcher and became Senior Research Scientist in 2006. He is the Scientific Leader of the INRIA Project Team AVIZ (see www.aviz.fr) that he founded in 2007 and that is well known worldwide in the domains of visualization and human-computer interaction. His main research areas are Visual Analytics, Information Visualization and Human Computer Interaction. Jean-Daniel Fekete was the General Chair of the IEEE VIS Conference in 2014, the first time it was held outside of the USA in Paris. He is an Associate Editor of the IEEE Transactions on Visualization and Computer Graphics (TVCG), Member of the IEEE Information Visualization Conference Steering Committee and of the EG EuroVis Steering Committee. During 2015, he is on Sabbatical at NYU and Harvard.

This talk is organized by Austin Beck