Friday January 18, 2019
Home Lead Story Facebook Ensu...

Facebook Ensuring Its AI System to be Equally Neutral For All

Facebook also announced that it was using AI to remove posts from its platform that involve hate speech, nudity, graphic violence, terrorist content, spam, fake accounts and suicide.

0
//
Facebook
An image showing a Facebook logo reflected in a person's eye. VOA

Facebook wants to ensure that its Artificial Intelligence (AI) system comes across as neutral to everyone so that nobody feels discriminated against in all the things that it does – from job recommendations to removal of posts that violate the policies of the social network.

The company has built a system called Fairness Flow that can measure for potential biases for or against particular groups of people, research scientist Isabel Kloumann was quoted as saying at Facebook’s F8 developer conference on Wednesday by CNET.

“We wanted to ensure jobs recommendations weren’t biased against some groups over others,” Kloumann said.

Facebook also announced that it was using AI to remove posts from its platform that involve hate speech, nudity, graphic violence, terrorist content, spam, fake accounts and suicide.
Facebook making its AI neutral for all. Pixabay

Facebook also announced that it was using AI to remove posts from its platform that involve hate speech, nudity, graphic violence, terrorist content, spam, fake accounts and suicide.

“We view AI as a foundational technology, and we’ve made deep investments in advancing the state of the art through scientist-directed research,” Facebook said in a statement on Wednesday.

Also Read: Twitter Asked 336 Million Users to Change Password as it Has Detected a Bug

At F8, its AI research and engineering teams shared a recent breakthrough: the teams successfully trained an image recognition system on a data set of 3.5 billion publicly available photos, using the hashtags on those photos in place of human annotations.

“We’ve already been able to leverage this work in production to improve our ability to identify content that violates our policies,” the statement added.

The announcements came even as the company finds itself in the midst of increased scrutiny over its data protection practices.

On the inaugural day of the two-day developer conference, Facebook CEO Mark Zuckerberg promised more steps to stop abuse of its services. (IANS)

 

Next Story

Facebook Kills 300 Fake Pages and Accounts Linked To Russia

The individuals handling these accounts primarily represented themselves as Ukrainian and they operated a variety of fake accounts while sharing local Ukrainian news stories

0
Facebook, social media
Facebook kills over 300 Russia-linked fake accounts, Pages, (VOA)

Social networking giant Facebook on Thursday killed over 300 fake Pages and accounts linked to Russia that it said was “engaging in coordinated inauthentic behaviour”, almost two years after the US presidential election.

The social media giant removed 364 Pages and accounts as part of a network that originated in Russia and operated in the Baltics, Central Asia, the Caucasus, and Central and Eastern European countries.

“The two operations we found originated in Russia, and one was active in a variety of countries while the other was specific to Ukraine.

“Despite their misrepresentations of their identities, we found that these Pages and accounts were linked to employees of Sputnik, a news agency based in Moscow, and that some of the Pages frequently posted about topics like anti-NATO sentiment, protest movements, and anti-corruption,” Nathaniel Gleicher, Head of Cybersecurity Policy, Facebook, wrote in a blog post.

Facebook, photos
This photograph taken on May 16, 2018, shows a figurine standing in front of the logo of social network Facebook on a cracked screen of a smartphone in Paris. VOA

About $135,000 were spent in ads, and paid for in euros, roubles and dollars, over a time span from October 2013 to now.

Around 790,000 accounts followed one or more of these Pages (now taken down) on the social media platform.

“We’re taking down these Pages and accounts based on their behaviour, not the content they post. In these cases, the people behind this activity coordinated with one another and used fake accounts to misrepresent themselves,” Gleicher added.

The pages hosted about 190 events, with the most recent scheduled for January 2019, and about 1,200 people expressing interest in them.

We cannot confirm whether any of these events actually occurred, the company added.

Facebook, data
Facebook staring at bigger problems this year, warns analyst. VOA

Separately, the social networking giant removed another 107 pages, groups and accounts as well as 41 Instagram accounts that were designed to look like they were operating from Ukraine, but were in fact part of a network that originated in Russia.

Also Read: Facebook Violated Cyber Security Law: Vietnam

The individuals handling these accounts primarily represented themselves as Ukrainian and they operated a variety of fake accounts while sharing local Ukrainian news stories on topics ranging from weather, protests, NATO to health conditions at schools.

“We identified some technical overlap with Russia-based activity we saw prior to the US midterm elections, including behaviour that shared characteristics with previous Internet Research Agency (IRA) activity,” the company said. (IANS)