$1.1M grant will fund reasearch on analyzing privacy-enhanced data

UNIVERSITY PARK, Pa. — The National Science Foundation (NSF) has awarded a $1.1 million grant to a multidisciplinary Penn State team to develop methods for statistical analysis of noisy and privacy-enhanced data.

Daniel Kifer, assistant professor of computer science and engineering, serves as principal investigator on the NSF grant. Stephen Matthews, associate professor of sociology, anthropology and demographics, and Tse-Chuan Yang, a research associate with the Social Sciences Research Institute, are co-PIs.

Kifer explained that information derived from confidential surveys that are analyzed must be “sanitized” so that individuals cannot be identified. To do so, "noise" is introduced into the original data for the individuals’ protection. For example, specific information about an individual who suffers from a rare medical condition may be omitted or perturbed in the privacy-enhanced data.

The problem, Kifer said, is when the noisy data is analyzed using traditional techniques, it may create false positives or misleading results.

The Penn State effort looks to create new techniques for extracting more accurate statistical information from these complicated, noisy datasets. The work will place special emphasis on analyses useful for social science research.

Media Contacts: 
Last Updated October 23, 2012