PrivaSee: Building an Automated Interface to Enhance Usability and User Comprehension of Privacy Policies

Abstract

This thesis presents PrivaSee, proof-of-concept demo interface for reading privacy policies to aid user decision making by improving readability, comprehension, and understanding of the policy in relation to other privacy policies in a corpus. It begins with a demonstration of the design of the interface based on user needs. I discuss the technical methods evaluated and used to build the interface. PrivaSee is then evaluated for its success in achieving its goals by presenting findings and analysis generated from user studies, showing preliminary success in improving reading comprehension and readability scores for the experimental user group.