Facebook Using New Rosetta AI to Moderate Memes and Content

By:  |  Category: Blog Tuesday, September 11th, 2018  |  No Comments
Rosetta AI

Facebook is currently facing tremendous pressure to moderate the content posted on their site. When Russian trolls, bullying, and hate speech fill up the feed, even the government get’s involved in monitoring the social media giant. But how do you look through every single image and comb through the data posted on the massive site?

Usually, computers rely on a method called Optical Character Recognition, or OCR, to classify what’s in pictures or videos, but because of Facebook’s enormous numbers — there are 2.2 billion people who use the social network each month — OCR isn’t enough. So Facebook needed a solution to work alongside their 20,000 social monitors with a bigger reach.

In response, Facebook has developed  an Artificial Intelligence (AI) system they call Rosetta AI  that utilizes machine learning to identify text in images and video and then translate it into data a machine can read. This is particularly helpful with memes-where both text and image come together.

Rosetta is now up and running at Facebook, extracting text from 1 billion images and video frames per day across both Facebook and Instagram.

Right now, its a little vague what Facebook is doing with all the data. Their blog states, “we are (now) able to process more than a billion public images per day through our system efficiently. The extracted text is used by downstream classifiers to immediately act upon policy-violating content or by product applications like photo search.”

So according to their developers, Rosetta will be used for basic features like photo search and screen readers,  but also things like figuring out what would be compelling to put in your News Feed, and, more importantly, figuring out which memes are just silly and which are actually spreading hate speech or inciting conflict.

This new AI tool will certainly help with public perception as Facebook faces scrutiny over content on its social network. Facebook has been accused of influencing elections in the US, to polarizing public opinion and worse, of helping to spur violence in Myanmar, Sri Lanka and India. Last month, the social network said it was taking action to stop “spread of hate” in Myanmar due to disinformation posted on Facebook. The social network said it’s removed 18 accounts and 52 pages associated with the Myanmar military because of ongoing ethnic violence against Rohingya Muslims.

In the last month, Facebook COO Sheryl Sandberg, along with Twitter CEO Jack Dorsey, appeared before an angry Congress over its lack of content moderation policies and security practices to keep users safe.

It’s exciting to see the potential for AI as a tool for improving business on a grand scale, and even more importantly, actually contributing a much needed hand of moderation to our escalating social climate.

If you need assistance with managed IT and Security Services give EnhancedTECH a call at 714-970-9330 or contact us at [email protected]


Leave a Comment
Read previous post:
Natural Language Bot
See How Spanish Telco Builds Digital Assistant Based on Natural Language Bot

Technology is now learning our language instead of the other way around. EnhancedTECH is here to help you leverage the...