MSc candidate Niharika Guntamukkala will defend their thesis “A Novel Automated Approach to Evaluate the Transparency of Privacy Policies” on May 4, 2016, at 2:00pm in Reynolds 219.

Title

A Novel Automated Approach to Evaluate the Transparency of Privacy Policies

Abstract

Research findings suggest that online privacy policies are often long, hard to understand and contain insufficient information. Therefore it is suggested that privacy policies should be transparent to help users make informed decisions about their personal information and avoid the risk of information misuse. In order to help users gauge the level of transparency, it would be beneficial to have an automated system that evaluates the transparency of a given privacy policy and provides useful feedback. In this thesis, we propose an automated transparency evaluation system to analyze the privacy policies based on their content and style. This tool gives a transparency score by analyzing the content factors completeness and unambiguity using machine-learning techniques and style factor readability using existing readability formulas, to provide feedback to the users and warn them about the risks associated with their information privacy. The system is able to achieve F-measure scores greater than 70% in classifying a privacy policy as ‘ most transparent ’ , ‘ moderately transparent ’ or ‘ least transparent ’ .

Advisor: Dr. Rozita Dara

Examination Committee: Dr. Deb Stacey, Dr. Dave Calvert (Chair)