Twitter is trying to figure out how to improve its algorithms

Twitter is trying to figure out how to improve its algorithms

The Responsible Machine Learning program will have the task of studying the algorithms of the social network to try to understand how to improve machine learning and keep it away from racial, gender or political bias

(photo: Omar Marques / SOPA Images / LightRocket via Getty Images) Twitter is trying to assess any damage unintentionally caused by its algorithms. Responsible Machine Learning is the initiative that will see a team of engineers, researchers and data scientists engaged in the study of the use that the social network makes of machine learning: it will be useful to understand how the platform's algorithms may or may not be predisposed to bias that have negative consequences on users.

To begin with, the algorithm that automatically crops images in tweets, uploads them to posts to preview them, will be tested and evaluated on racial and gender bias. In fact, many Twitter users have reported that too often automatic image previews seem to favor the faces of white people over black ones. This initially prompted the company to test a system that shows the full image and no longer a cropped preview. However, since this solution would not solve the problem related to the alleged algorithmic bias, Twitter has decided to act and to solve this problem.

The Rml team will also have the task of analyzing the recommendations of the history which, at as reported by some users, they differ according to racial and gender subgroups; it will then analyze the recommendations on the contents also based on political ideologies and expanding the analysis on more countries.

Twitter, however, will not work alone, sharing the results of its analyzes and asking for feedback from its users. The social network is confident about the work that this new team will go to accomplish as, as stated in the company's blog post, the results obtained "may not always translate into visible changes to the product".


LOL - 24 hours ago

Facebook mistakenly removed the page of the French town of Bitche

adsJSCode ("nativeADV1", [[2,1]], "true", "1", "native", "read-more", "1"); Social Network - Apr 15

What is Frank, the new conservative social network you can't swear on

adsJSCode ("nativeADV2", [[2,1]], "true", "2", "native", "read-more", "2"); Tech - 15 Apr

The Italian Privacy Guarantor investigates the most controversial facial recognition company in the world

Topics

Artificial intelligence Social Network Twitter globalData.fldTopic = "Artificial intelligence , Social Network, Twitter "

You may also be interested in



This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.




Powered by Blogger.