In an attempt to get an even better understanding of its members and their data, Facebook has launched a new research group focused on using artificial intelligence to make more precise judgments about a user’s text, photos, and future behavior.
The approach — known as deep learning — uses simulated networks of brain cells to process data. By applying this method, Facebook might be able to show content that is more relevant to members’ interests and improve ad targeting.
Mike Schroepfer, the company’s chief technology officer, said that one obvious place to use deep learning is to improve the News Feed. Facebook already uses “conventional machine learning techniques” to cut back the 1,500 updates that the average Facebook member could see down to 30 to 60 that are judged to be most likely to be important to them.
Conventional forms of machine learning are much slower because before data can be fed in, “experts” must manually choose which features the software should pay attention to. Deep learning systems can learn with much less human intervention because they can figure out for themselves which features are most useful to understanding it.
This isn’t the first time a company has relied on deep learning — Google and Microsoft have both used the approach in the past year. In fact, last year Google created software that taught itself to recognize cats and other objects by reviewing stills from YouTube videos. This technology was later used to minimize the error rate of Google’s voice recognition services.
“The data set is increasing in size, people are getting more friends, and with the advent of mobile, people are online more frequently,” said Schroepfer. “It’s not that I look at my news feed once at the end of the day; I constantly pull out my phone while I’m waiting for my friend, or I’m at the coffee shop.”
If this method can truly improve Facebook’s ability to predict actions and provide more relevant content and ads, the social network could become a much more lucrative platform for marketers. The company’s AI group will work on both applications that can help its products as well as on more general research around the topic that will be made public.