In the current age of big data, we are continually creating new data which is analyzed by various platforms to improve service and users’ experience. Often this data is sensitive and confidential in nature, which raises obvious security and privacy concerns while storing and analyzing such data. In this talk, I will review fundamental challenges in providing robust security and privacy guarantee for large scale data management and analysis, and give an overview of my contributions towards addressing these challenges in the framework of differential privacy. Further, I will exemplify key ideas through my recent work on spectral sparsification of graphs. I will conclude with my vision of robust and private scalable data analysis.
For more information click here.