Jenni Reuben Shanthamoorthy
I am interested in problems concerning privacy challenges in data analyses. In this broad context, my focus is on the processing of interconnected data models such as graphs. Randomization is one of intuitive approach for distorting the output of an analysis to achieve a meaningful privacy guarantee. Differential privacy is a privacy model that provides a formal definition for such a privacy guarantee. With the increase in the use of graph in representing social networks, financial transactions, etc., many research endeavors has turned to apply this model for releasing graph statistics. However, the challenge is to release an accurate graph statistics yet achieving a meaningful privacy guarantee. To this end we propose a variant definition of graph dataset neighbors for formalizing the differential privacy definition for edge-labeled graphs. In my research, I am interested in empirically evaluating the privacy/utility implications of this definition versus other graph dataset neighbor definition for a range of graph statistics.
Teaching activities at KAU for which I am was a co-teacher:
- Spring 2018: DVGB07 - C#.NET.
- Spring 2018: DVGC25 - Supervision of Bachelor Thesis. Pontus Anttila, "Mot effektiv identifiering och insamling av brutna länkar med hjälp av en spindel"
I hold a dual master degree in Computer Science with main focus in information security, from KTH (Kungliga Tekniska Högskolan), Sweden and Aalto University, Finland. After which, I have worked as a software engineer for 51/2 years for industry leaders such as Verizon, Citi group and Symantec before starting my PhD endeavor.