Can Mouse Movements Reveal Your Behavior?

We already know that the browser you’re using and how you’re using it can reveal a lot about your online behavior through browser fingerprinting, but did you ever consider how much can mouse movements reveal your behavior?

According to the computer scientists at the University of Luxembourg and international partners, this is just the case.

In two recently published research papers, they revealed how mouse movements can reveal some rather interesting things about the user, including sensitive information such as their age, gender and more.

Their papers have been shown at the 6th ACM SIGIR Conference on Human Information Interaction and Retrieval.

Key Findings

One of the authors of the papers, Professor Luis Leiva explained the key findings:

We have demonstrated how straightforward it is to capture behavioral data about the users at scale, by unobtrusively tracking their mouse cursor movements, and predict user’s demographics information with reasonable accuracy using five lines of code. For years, recording mouse movements on websites has been easy, however, to analyze them one would need advanced expertise in computer science and machine learning. Today, there are many libraries and frameworks that allows anyone with a minimum of programming knowledge to create rather sophisticated classifiers. This raises new privacy issues and users do not have an easy opt-out mechanism.

What are the Applications for Search Engines and Webmasters?

While the idea of tracking mouse motion trajectory raises some clear privacy alarms, it actually has practical applications for search engines and webmasters.

According to the co-author of the two papers, Dr. Ioaniss Aparakis:

When you search for something at Google or Bing, your mouse movements are sending a tiny signal that the search engine indicating if you are interested or not in the content you have been shown. As mouse tracking may have privacy issues, we investigated the possibility of recording only a small part of the whole movement trajectory and see if we can still infer how people make choices in web search.

Three Mouse Cursor Movements Scenarios

The researchers specifically analyzed three different scenarios in which users had to choose on their search engines:

  1. Noticing an advertisement;
  2. Abandoning a page;
  3. Becoming frustrated.

Here’s what they found out:

  1. If the user noticed an ad, they will signal this with initial mouse movements;
  2. In case of abandoning a page, it’s the last mouse movements that tells if the user is going to leave, even before clicking on anything;
  3. Finally, when becoming frustrated, it’s the middle part of a cursor movement trajectory that provides the strongest signals about user behavior.

The researchers discovered that they can predict these three scenarios with just two or three seconds of mouse movement and have concluded that search engines can get useful information and improve their services without violating user’s privacy by tracking only interesting parts of mouse movement.

According to Professor Leiva:

By efficiently recording the right amount of movement data, we can save valuable bandwidth and storage, respect the user’s privacy and increase the speed at which machine learning models can be trained and deployed. Considering the web-scale, doing so will have a net benefit on our environment.

Conclusion

Finally, the researchers have also developed a method in the form of a web browser extension called MouseFaker to prevent mouse tracking. The extension distorts mouse coordinates in real-time.

Speaking about the extension, Professor Leiva said:

It is inspired by recent research in adversial machine learning and has been implemented as a web browser application, so that anyone can benefit from this work in practice.

The extension is available on Github.