Guadalupe Canahuate and Michael Gibas, CSE Ph.D candidates, received awards for their research projects at the 4th annual Midwest Database Research Symposium. Guadalupe and Michael are advisees of Hakan Ferhatosmanoglu in his Database Research Lab. This pair of awards were chosen from approximately 50 posters presented.
Guadalupe's Update Conscious Bitmap Indices poster presentation, co-authored by Michael Gibas and Hakan Ferhatosmanoglu, won the award for the most novel work. Bitmap indexes have been successfully applied towards fast query processing over static data sets. However, their use for dynamic data domains has been limited due to the time required to update the index to reflect changes in the data. This work significantly enhances bitmap indexes by providing a technique to minimally modify the index to accurately reflect appended data.
A General Framework for Modeling and Processing Optimization Queries, Michael's entry, (co-authored by Ning Zheng and Hakan Ferhatosmanoglu) was voted the most promising work. This work provides a methodology to I/O-optimally discover the best points in a database according to some arbitrary convex function while incorporating arbitrary convex problem constraints over arbitrary access structures built on convex partitions. This provides a unified framework to efficiently answer a wide range of ranked queries, which enhances the effectiveness of scientific discovery.
Midwest Database Research Symposium strives strengthen the ties between the database researchers at universities in the region and to encourage future discussions and interactions through poster presentations and networking. This year's symposium was held at Purdue University and included participants from more than 15 universities, including some of the nation's top database research groups. More information regarding the symposium can be found at http://dais.cs.uiuc.edu/mwdbrs/.
Congratulations to all of CSE's participants for a successful symposium and their innovative research.