Privacy Beyond Privacy Settings: A Twitter Case Study

We use machine learning techniques to build a model that is capable of looking at user’s Twitter data and determining their political tendency. This approach demonstrates how malicious actors could build such a model and use that model in disinformation campaign. This work is specially critical in an election year and intends to demonstrate how fragile democracy is.

If users knew how their Twitter activity would be perceived then they might be more careful with their activities. We have a built a privacy tool — a website to help users know exactly that! Our website enables users to learn about their political tendency scores based on their Twitter activities!