Monday, February 6, 2017

Facebook makes an agreement with 8 the French media to create an algorithm anti buffaloes – Agora Magazine

Eight French media, including Le Monde, will work with Facebook to “reduce the presence of false information on the social network” and create a sort of super-algorithm anti-buffaloes.

“In the concrete – a report in the newspaper of Paris on its web site – the project consists in deploying soon in France, a device similar to the one that was put in place in December in the United States, with the participation of the five media (Abc News, Ap, FactCheck.org, Politifact and Snopes) and should soon be launched in Germany, with the drafting of Correctiv. In France, In addition to Le Monde, the media partners are the agence France-Press, Bfm Tv, France Televisions, France Medias Monde, L’express, Libération and 20 Minutes. After France, Facebook plans to advance the development in other countries”.

The device, explains the daily, will allow users, thanks to a new category of alert, to detect a piece of news which is suspected to be false. Links reported end up collected in a portal they have access to the media partners, who can verify the information. If the two media determine that the content is fake, the news will appear to users with a token red by mentioning that the fact-checker he dispute the truthfulness. An alert window will also open when users share the link, and the operation will have no effect on the algorithm that the program the dissemination of the contents.

Facebook “trains” the algorithm to recognize the fake news

January 31, Facebook announced in a blogpost the official new changes to the algorithm with the aim to "identify and reward authentic content" and thus continue in the fight against fake news. As you can read on The Print, in order to obtain these results, Facebook will use a large sample of pages known to disseminate news, offer an overview or false, by training the algorithm to recognize the particular features and to penalize the fanpage that spread the content of this type. With the agreement concluded with the French media, this “training” is much more effective and, as explained in the director of Le Monde, Jerome Fenoglio, “per the first time, it will be possible to intervene on an algorithm when the content poses a problem for publishers. It is precisely this element which has convinced us”. In the battle of the fake news, Facebook has also decided to reserve more and more attentio n to the events and to post that they are getting hit, trying to identify them and address them through an algorithm of the predictor, which will analyze in real time the number of likes, comments and shares that the post are generating, in order to show at the top of the newsfeed of the users of the most successful, increasing timely visibility.

Immigrant syrian port Facebook in court in Germany

In Germany, meanwhile, a syrian refugee has brought Facebook to court to put an end to the spread of fake news on terrorism and on immigration, containing a picture of her. As reported in The Paper, Anas Modamani, who fled from the syrian civil war in 2015, and took refuge in Germany, became famous when he took a selfie with Angela Merkel during a visit of the German chancellor at the center of reception of immigrants in Berlin. That picture in the short time he has been around the world and the selfie with Merkel of the twenty-year-old, who now lives and works in Germany, is often used on Facebook by those who want to attack immigrants.

Support your newspaper Agora Magazine, our newspapers do not have public funding. Thanks Agora Space Editor

LikeTweet

No comments:

Post a Comment