Public Data, Public Good, Personal Information: Techniques and Challenges

This event is part of our Social and Decision Analytics Seminar Series. 

Speaker: Madeline Pickens, Data Scientist at the Urban Institute

Abstract: Public policy researchers rely heavily on publicly released datasets and statistics. However, even for datasets that have been scrubbed of Personally Identifiable Information (PII), there remains a risk that individuals in the data could be re-identified by malicious actors. This risk continues to increase as access to substantial computational power and auxiliary datasets becomes more common. To maximize public good, data stewards have a responsibility to maintain the privacy of individuals in their data while simultaneously making the released information as useful as possible to data users. This talk will include a high-level overview of available quantitative techniques used to obscure personal information, challenges in their implementation, and methods for measuring the tradeoff between data privacy and utility. Finally, several ongoing projects at the Urban Institute will be presented to illustrate these concepts in practice.

Join the event here.