Ever try to read a Privacy Policy? It seems that they're there, but they aren't meant to be read. Privacy Policies are important: they both enable and restrict a company to use information about you. It shouldn't be so hard to understand such an important part of our digital lives.

We attacked the problem by leveraging the power of Natural Language Processing. We hunt for clues in any Privacy Policy to outline what the policy is actually trying to say, and cutting out all the fluff that makes it hard to comprehend.

A Naive Bayes Classifier is run through the entirety of the Privacy Policy we're analysing, calculating probabilities that the Policy enables certain data use and collection such as targeted advertising. We also try and find hints of 3rd party involvement with your data, exposing the companies who refuse to respect the privacy of their users.

We hope to improve our algorithms to be able to generate accurate summaries of privacy policies that are generated on the fly, and actually human readable!

Share this project: