Facebook’s fake news detection systems extends to photos and videos

The fight against false information on Facebook, which has reached the highest international institutions, continues. According to the company’s product manager, Antonia Woodford, the tools for the control of the veracity of the publications in the social network now extend to the photographs and videos in 17 countries, the country said.

Information poisoning through the social network for the spread of lies led the company to establish a system of human teams that review the publications most likely to hide an intention to further ravel the network and intelligence programs artificially sweeping data permanently.

This system will now be applied to videos and photographs, two formats on the rise in all platforms and of great capacity to propagate falsehoods.

The mechanical learning model looks for signals and alert messages from Facebook users to, according to Woodford, “identify potential false content”. This is added to the equipment of data contrasters that, for the new function, have been prepared in the evaluation of images by means of “visual verification techniques” that compare the images on the same object and analyze the metadata (the hidden information that Includes any digital file).

The offensive of Facebook in this field responds to the rise of the diffusion of audiovisual material through the network, because of the high capacity of influence and penetration of this type of messages and the ease for them to be altered.