From Data to Information

Abstract

Phone: 301.405.6722 Whether the problem is making sure that any g iven geographical location is represented once and only once in a database, figur ing out who’s who in a corporation based on email cor respondence, discover ing the social structure among a group of dolphins, or mapping how genes regulate each other in human cells, Lise Getoor and her strategy of graph identification can help suggest a solution. Getoor, a member of UMIACS and an assistant professor of computer science, descr ibes graph identification as a process that takes data, which is generally jumbled, noisy, and redundant, and yields information, which is tidy, log ically organized, and wellannotated.