Thousands of patents for reading and analysing user data and preferences have been filed by Facebook since 2012. For example, one of them involves using a front camera to analyse the user’s facial reaction to posts, and another involves using machine learning algorithms to predict life-changing events in the lives of its users. Here are seven such patents from the recent past that The New York Times thought was worth reviewing in a recent story:
Reading the user’s relationships
One patent talks about predicting if the user is in a romantic relationship based on the number of times they visit another user’s page, the number of people in their profile picture, and the percentage of friends they have who are of a different gender.
Classifying the user’s personality
Another patent attempts to understand the user’s personality based on their posts and messages to serve them more relevant stories or ads. It attempts to judge the user’s degree of extroversion and emotional stability.
Predicting the user’s future
This patent proposes to predict major events in the user’s life like marriage, graduation, childbirth and death by analysing credit card transactions and location information, in addition to the user’s posts and messages.
Identifying the user’s camera
There’s one patent that talks about identifying the user’s camera by creating a unique camera ‘signature’ for every user, using faulty pixels or scratch marks on the lens. By knowing who uploads pictures taken through these cameras, the relationship or ‘affinity’ between the uploader and the camera user can be guessed.
Listening to the user’s environment
Another patent proposes listening in on the user’s surroundings to understand which TV shows are watched and which ads are muted. It also proposes guessing the name of the TV show by using the electrical interference pattern caused by the television’s power cable.
Tracking the user’s routine
One more patent application talks about tracking the user’s weekly routine and sending notifications of any deviations in it to other users. It also talks about determining the user’s home address by reading location information in the middle of the night.
Understanding the user’s habits
This patent proposes using the user’s location along with their friends’ location to determine whom the user socialises most with. It also proposes tracking the number of hours the phone is stationary to count sleeping hours.
According to The New York Times, Allen Lo, a Facebook Vice President said in an email, “Most of the technology outlined in these patents has not been included in any of our products, and never will be.”
Though Facebook iterates that its patent applications are never indicative of an upcoming product plan, it’s hard not to look at the social media company with a jaundiced eye, especially after the recent data scandal the company and its CEO were embroiled in.
Google has also created an AI algorithm that accurately predicted the death of a woman suffering from breast cancer. Earlier, it developed an algorithm that found two exoplanets.