Over the last year, a team of as many as 260 workers in Hyderabad, India, had been going through millions of Facebook Inc. photos, contents and status updates posted since 2014, which had amped up further privacy concerns for the users of the social networking mogul.
According to a Facebook spokesman, the contract workers on Hyderabad had been categorizing the items according to five categories such as subject, occasion and intention of the post, and the manual content labelling had been aimed at understanding the changing behaviors of the Facebook users, which could help the company come up with new features, increase usage and surge ad revenues.
Nevertheless, the Wipro Ltd contract workers, to whom the Facebook Inc. had bequeathed the project, had said in terms of anonymity that they had been categorizing the posts on the basis of subject, occasion and intention.
As the labelling effort had not been disclosed, the content marking program could add further strains into the grief-sicken Facebook Inc. which had currently been facing dozens of lawsuits worldwide over potential privacy issues.
However, adding that the labelling effort had always been a core part for AI-based feature development on Facebook, the director of product management for AI at Facebook, Nipun Mathur said, “It’s a core part of what you need. I don’t see the need going away”