Facebook Cares About Your Mood

Facebook seems to leave no stone unturned when it comes to attempts to keep its customers glued to the application. A craze among youngsters, Facebook has transcended the barriers of age and time. As of March 2017, there are over 1.94 billion monthly and 1.28 billion daily active Facebook users worldwide. With five new profiles created every second, the customer or audience base of the social media giant is ever expanding. Every 60 seconds on Facebook finds 510,000 comments posted, 293,000 statuses updated, and 136,000 photos uploaded. Facebook’s patents span different technologies, from network bandwidth management, user authentication and privacy controls, to pattern detection and customized content delivery. Facebook has been successfully patenting some of its proprietary and popular features, such as the technology behind its News Feed, Timeline, in-network messaging, and digital media tagging features.

To an extent, at any given second 20,000 people are active on Facebook. But from Facebook’s perspective, the biggest challenge is to keep the attention of its viewers intact, preventing them from clicking their way out to other apps and websites.

Enters the Boredom Detector! In a recent patent application, filed in Jan 2017 and published in May 2017, Facebook describes a system that determines if a user is dissatisfied, uninterested, or bored with the content displayed in feeds, and quickly reshuffles the content to tempt one to stay. The system measures the length of time from when a user starts scrolling through the feed to the time that user stops scrolling. The longer one goes without finding something that makes the thumb stop, the likelier it is that the user is bored. To make sure its assumption is correct Facebook proposes to track the eye position using a device’s camera ad check if a user loses focus on the display screen for a certain duration. If identified, a fresh look is taken at all the items on display and a re-ranking is conducted in real time. The interface then makes room for any new items that have been posted since the scrolling began.

 

Img Src: Figs 3 and 4 of US20170126825

Image Source: Figs 3 and 4 of US20170126825

In another similar disclosure, Facebook proposes a system for Emotion Detection-based content delivery. Simply put, this patent application suggests using the webcam or smartphone camera in real-time as the user browses, to understand change in facial expressions and assess the user’s feelings across different types of content. For instance, if a user smiles while looking at pictures of a friend, Facebook’s algorithm will make note of that and cater more pictures of that friend in the user’s feed. If a user looks away from the screen when a video of a particular type is played, Facebook will stop showing similar videos. In a similar way, if the user happens to watch an advertisement with amusement, Facebook could choose to target that user with more similar adverts to entice the user on the service for longer.

This patent application was submitted in February 2014 and published in August 2015. While Facebook already has its own ways to self-curate the content on News Feeds through unfollowing and other methods, the proposed facial emotion detection system might be a significant move to expand and automate this process.

Imahe Source: Figs 3 and 4 of US20150242679

Image Source: Figs 3 and 4 of US20150242679

Pushing technological boundaries despite controversies is something Facebook has been famous for. In a related patent application titled ‘Augmenting text messages with emotion information’, filed in November 2015, Facebook details how emotions of a user can be predicted based on “characteristics of a keyboard input”, such as speed of typing and pressure of how one taps the keys. The patent further proposes that features can be added to texts to resonate with the feelings of the sender. Another patent proposal based on facial recognition focuses on the generation of emojis based on the facial expression of a user. For example, the user can take a selfie or a photo and the system will produce an emoji expressing the same emotion. This application was filed in November 2015 and is titled ‘Systems and methods for dynamically generating emojis based on image analysis of facial features’.

Image Source: Figs 3A, 3B of US20170147202 and Fig. 4 of US20170140214

Image Source: Figs 3A, 3B of US20170147202 and Fig. 4 of US20170140214

Whether these newly proposed features will get a Like or a Dislike from users is still debatable. If implemented, these inventions may be premised as invading the privacy of users during their most personal moments.

Earlier this year, Facebook engaged in an experiment to track the emotional state of its teenage users through their posts and Facebook activity. The data gathered was used to showcase customized ads. In 2014, Facebook also conducted a large-scale experiment where it adjusted the feeds of around 700,000 Facebook users and monitored their posts and comments to understand if social media activity could manipulate people’s emotions. In another similar study, Facebook collaborated in a research with Cornell and the University of California by filtering users’ news feeds, including the flow of comments, videos, pictures and web links posted by others. One test reduced users’ exposure to their friends’ “positive emotional content”, resulting in fewer positive posts of their own. Another test reduced exposure to “negative emotional content”. The study concluded that emotions expressed by a user’s friends over social media affected their emotional state of mind.

Facebook’s recent patent applications might fuel speculation and add worries of privacy invasion among users. Facebook officials however say that the filing for the webcam detection patent is purely exploratory, and patent filings aren’t a sure indicator of a company’s product roadmap. While Facebook’s popularity and ubiquity was so far not significantly affected by past allegations of privacy violations, the consequence of privacy invasion by secretly taking photos of users without their consent is something the company should seriously consider, if at all it plans to put the theories to practice.

(Featured image is for representational purpose alone. Image has been sourced from https://pixabay.com/p-1624400/?no_redirec)

Subhasri Das
Subhasri Das

Subhasri is a technocrat who enjoys reading between the lines of patents to understand their hidden value.


Post a Comment

Your email address will not be published. Required fields are marked *