Geetha Jagannathan

Research Interests

My research lies in the general area of Trustworthy Computing, with an emphasis on privacy-preserving data analysis and secure methods for distributed computation. The purpose of my research is to develop algorithms, protocols and theories for preserving the privacy of individuals and institutions when their data is released for public use or when their data is used in the computation of aggregate structures. My interests are currently focussed on practical methods for: (i) constructing utility efficient data mining techniques from differentially-private summaries, (ii) differentially-private anonymization of graphs such as social networks, (iii) differentially-private release of time series and (iv) differential privacy for distributed data. Much of my research involves creating new machine learning/data mining algorithms that preserve privacy.