Jaspreet Bhatia

4208 Wean Hall, Carnegie Mellon University

I am a fourth year Ph.D. student at the Institute for Software Research at Carnegie Mellon University. I am fortunate to be advised by Dr. Travis Breaux.

I am an interdisciplinary researcher. My research lies at the intersection of natural language processing, machine learning, and quantitative and qualitative research methods. I study how users interact with software, what factors contribute to their perceptions of privacy and how can we develop technologies to help users make informed decisions about using software systems. More specifically, my research work focuses on developing techniques to extract and analyze privacy requirements using crowdsourcing, natural language processing, deep learning, and qualitative data analysis. I use quantitative data analysis techniques to determine how users perceive their interactions with a website, and have developed an empirical framework to measure privacy risk.

During my PhD so far, I have developed a hybridized task re-composition framework, that semi-automatically extracts privacy goals that describe a website's data practices. I have also worked on developing a theory of vagueness and privacy risk perception, with our collaborators, Prof. Joel Reidenberg and Dr. Thomas Nortan.

I have developed an empirically validated framework for understanding and measuring perceived privacy risk.

I am currently working on developing a deep learning framework to extract and analyze data practices described in privacy policies, in the form of semantic frames using LSTMs. These frames help store the semantic information extracted from the data practice in a structured format, enable question-answering about the privacy policy, and help users and regulators better understand the website's data practices.


30 May 2018 Our paper on grounded analysis of semantic roles in privacy goals has been accepted to RE 2018.
28 February 2018 I presented our privacy risk work at PrivacyCon 2018.
9 November 2017 I presented (and passed!) my thesis proposal titled Ambiguity in Privacy Policies and Perceived Privacy Risk on 9th November 2017 at Carnegie Mellon University.
6 September 2017 I presented our paper describing a case study of data purposes in privacy policies at RE 2017.
May 2017 Our papers on automated hypernymy extraction and on the case study on data purposes have been accepted to RE 2017.
March 2017 I am on the program committee for AIRE 2017 workshop. Please consider submitting a paper! .
Feb 2017 Our work on developing a framework for understanding and measuring privacy risk has been accepted for discussion at Privacy Law Scholars Conference 2017.
Dec 2016 Our TOSEM paper on extracting privacy goals using hybridized task re-composition framework has been invited for presentation at ICSE 2017.
16 Nov 2016 Our paper on ambiguity in privacy policies and its impact on regulations received an Honorable Mention for Privacy Papers for Policymakers (Press Release).
18 Oct 2016 Presented our work on perceived privacy risk at the C3E workshop at Georgia Institute of Technology, Atlanta.
16 Sep 2016 Our paper on vagueness and risk perception was nominated for best paper award at RE'16 in Beijing.
15 Sep 2016 I will be presenting our paper on vagueness and risk perception at RE 2016, in Beijing (China) on 15 September 2016.
Summer 2016 Mentored an (awesome!) undergraduate REU student, Morgan Evans. Developed a technique to automatically identify and extract information type hyponymy relationships in privacy policies.
21 Mar 2016 Our paper on extracting privacy goals using hybridized task re-composition framework has been accepted for publication to TOSEM journal first edition.
16 Oct 2015 Joel Reidenberg and I, presented our work on comparing privacy policy ambiguity and its impact on regulation at the Contracting over Privacy Workshop, at University of Chicago.
Aug 2015 Presented our paper on automatically extracting information types from crowsourced privacy policy annotations at RELAW workshop held at RE 2015, in Ottawa (Canada).

CV  |  Google Scholar  |  LinkedIn