Story posted February 07, 2008
A new research and teaching laboratory is in the works for Bowdoin's computer science department, thanks in part to a grant from the National Science Foundation. Assistant Professor of Computer Science Laura Toma recently received the three-year $264,811 grant to continue her work developing scalable algorithms for realistic terrain processing in geographic information systems (GIS) applications.
A major component of the grant is development of a GIS laboratory at Bowdoin, to fully engage undergraduate students in contemporary research spanning theoretical algorithms and algorithmic engineering.
Toma has already worked extensively with undergraduate researchers at Bowdoin to help her develop efficient solutions for a number of fundamental problems and implement them into modules used by the GIS community, most notably, for open-source free software. Several of her students have presented their
Increasing support for student research fellowships is a key component of The Bowdoin Campaign, which ends in 2009.
"Terrain processing research is very motivating for students because it works with real applications," notes Toma. "It's something they can visualize."
Essentially, Toma and her colleagues are creating algorithms that manage and reduce computation time for the massive amounts of geographic data used in GIS applications. The scope and power of those geographic information systems are continually increasing as massive amounts of high-resolution data become available, for example, from remote sensing technology.
"Many GIS applications are not scalable under this volume of data, as they are designed only with CPU performance in mind," says Toma. "The bottleneck, however, is typically the data transfer between memory and disk. Research into the area of I/O-efficient algorithms has shown that performance can get greatly improved by designing algorithms that optimize both CPU and disk I/O.
"The goal is to build solutions that time very fast, that process huge data from satellites to make them accessible to the GIS community in the entire world," says Toma.
Toma notes that an open-source focus "is really in synch with the Common Good. I want my research, and my students' work, to contribute to a bigger, better world. I hope this research helps to bridge the divides between theoretical and practical computer science—and between computer science and geo-sciences—by developing algorithms and engineering GIS tools that are optimal, simple, practical and robust."