carlos med

Carlos Scheidegger

AT&T Labs

Ph.D. in Computing, 2009
Dissertation Title: Provenance of Exploratory Tasks in Scientific Visualization: Management and Applications
Advisor: Cláudio T. Silva

Carlos Scheidegger is the focus of this quarter's SCI Alumni Highlight. Carlos studied with Cláudio T. Silva at the SCI Institute, where he received his Ph.D. in 2009. His dissertation focused on the management of computational experiments, and how an appropriately designed software infrastructure can help analysts better understand and explore their problem domain. The paper describing his research was chosen as Best Paper at IEEE Visualization 2007. Key parts of his dissertation were implemented in VisTrails, an open-source scientific workflow system that tries to explore the research questions that arise when provenance capturing infrastructure is employed in real-life situations.

After joining AT&T Labs in 2009, his research interests shifted to the practical challenges that arise when refining visual data analysis software (for improved scaling, simplified writing, and efficient editing). The experience of building computational experiments has made him acutely aware of how software bugs tend to creep into research (some of it his own!), and how data reproducibility depends on the very foundation from which visualization software is built. This led him to research verification techniques in collaboration with his Ph.D. advisor and other research groups around the world. Some of this research has been published at IEEE Visualization 2009 and IEEE TVCG.

More recently, Carlos has become interested in how to bring visualization software to the internet, and how to rethink the software stack for programming graphics APIs. This research effort is currently very much in progress, but is currently being carried out in Facet, a domain-specific embedded language for large-scale visualization that attempts to bridge the gap between the very disparate GPU model and a modern web programming environment. If you have WebGL installed, you can try this Demo of Facet.

facet edge bundling
Facet, a domain-specific embedded language for WebGL graphics. Edge Bundling for Visualizing Large Graphs*

* E.R. Gansner, Y. Hu, S. North, C. Scheidegger. "Multilevel agglomerative edge bundling for visualizing large graphs" In Proceedings of the 2011 IEEE Pacific Visualization Symposium (PacificVis), pp. 187--194, 2011.