Privacy Policy Analyzer

A tool to get the essence of lengthy privacy terms in seconds.

How did you come up with the idea?

A few years ago Big Data was a just buzzword. Most companies were interested in using big data, but few had the hardware and software to actually do it. In just a few years things have changed dramatically, mainly due to affordable cloud computing and specialized software. Today, virtually any company can create extremely accurate user profiles by analyzing big data, which should be gathered with the consent of the user. And with Internet of Things around the corner this is merely the beginning.
Looking at these trends but also at the growing concern towards online privacy, we realized that users need to know more than ever what they are accepting when they tick that little box: “I agree with the Privacy Policy”. This is why we have created a Privacy Policy Analyzer, which automatically scans and detects possible issues with the Privacy Policy of a web service. This tool makes the web more transparent and simpler for users to understand.

As featured in...

  • AVG Web TuneUp

  • AVG Do Not Track

How does it work?

To assess the Privacy Policy of a web service we perform text analysis using natural language processing technologies that detect if any Personal Identifiable Information (PII) is being collected.
Then, this proprietary AI learning engine gets some help from human privacy experts who annotate a set of selected privacy policies (creating a training set), and marks fragments that are relevant to certain categories in the catalogue.
After the privacy expert completes annotation and the training set is prepared, the patent pending linguistic technology creates a model of the training set that helps the machine identify relevant fragments in unknown documents – such as third party privacy policies.

How does it look?

policy_miner_screen