Privacy Policy Summarization and Detection of Unfair Clauses

Abstract

ABSTRACT
Less than 1% of all users read the privacy policies (or terms and conditions) that accompany software products. This can be attributed to their low level of readability, length, and complexity. Hence, this allows corporations to sneak in clauses which may compromise a user’s privacy, allowing them to use their data for monetary benefit. For example, corporations may sell user data to third parties for targeted advertising. To tackle this, we propose a tool that will help users make a more informed decision before accepting such policies. We aim to improve the performance of the existing clause categorization algorithms by utilizing state-of-the-art deep learning models instead of the deprecated and proprietary Google Prediction API, and Support Vector Machines used in previous work. We propose a browser extension that will process a web page containing such policies and display the clauses related to common privacy issues, unfair clauses and links to related lawsuits.

Keywords: privacy policy, deep neural networks, categorization, unfair clauses, browser extension, lawsuits

Cite this Article: Jyoti Kulkarni, Divyank Agarwal, Pranav Barve, Rohit Awate, Stuti Biyani. Privacy Policy Summarization and Detection of Unfair Clauses. International Journal of Embedded Systems and Emerging Technologies. 2020; 6(1): 35–41p.