On March 23rd, Boston University held the\u00a0BU Science Day 2011 Symposium, with participation from over 180 graduate students and postdoctoral fellows. Out of a total of 60 considered, two posters were selected for the inaugural Hariri BU Science Day Awards,\u00a0\u00a0exemplifying\u00a0the power from adopting a computational approach in science and engineering.<\/p>\n
The titles and\u00a0abstracts of (and links to) the winning posters are included below.<\/p>\n
Dora Erdos, Vatche Ishakian, Andrei Lapets, Azer Bestavros, and Evimaria Terzi In many networks, (e.g., RSS feeds, blogs, sensor networks, ad-hoc networks)\u00a0 nodes blindly relay information they receive to neighbors. This uncoordinated data dissemination often results in significant, yet unnecessary, communication and processing overheads and reduce the utility of the network. To alleviate the negative impacts of information multiplicity, we propose that a subset of nodes (selected at key positions in the network) carry out additional information de-duplication functionality.\u00a0 We refer to such nodes as filters. We formally define the Filter Placement problem as a combinatorial optimization problem, and study its computational complexity for different types of graphs. We also present polynomial-time approximation algorithms for the problem. Our experimental results indicate that less than a handful of filters are enough to alleviate more than 95% of the redundant information.<\/p>\n
Dept of Electrical and Computer Engineering, Boston University<\/em><\/p>\n