Twitter admits algorithm bias for right-wing politicians and media | Twitter


Twitter has admitted that it amplifies more tweets from right-wing politicians and media than content from left-wing sources.

The social media platform reviewed tweets from elected officials from seven countries – the UK, US, Canada, France, Germany, Spain and Japan. He also investigated whether the political content of news outlets was amplified on Twitter, focusing primarily on American news sources such as Fox News, The New York Times and BuzzFeed.

The study compared Twitter’s “Home” timeline – the default way its 200 million users receive tweets, in which an algorithm adapts what users see – with the traditional timeline where tweets tweet them. newer ones are ranked first.

The study found that in six out of seven countries except Germany, tweets from right-wing politicians received more algorithm boost than those from the left; the right-wing media outlets were more amplified than those on the left; and generally politicians’ tweets were amplified more by an algorithmic timeline than by the chronological timeline.

According to a 27-page research paper, Twitter found a “statistically significant difference in favor of the political right” in all countries except Germany. For research purposes, a value of 0% meant tweets reached the same number of users on the algorithmically customized timeline as on its chronological counterpart, while a value of 100% meant tweets reached double. of the scope. On this basis, the largest gap between right and left was in Canada (Liberals 43%; Conservatives 167%), followed by the UK (Labor 112%; Conservatives 176%). Even excluding senior government officials, the results were similar, according to the document.

Twitter said it was unclear why its hosting timeline was producing these results and indicated that it may now need to tweak its algorithm. A blog post by Rumman Chowdhury, director of software engineering at Twitter, and Luca Belli, researcher on Twitter, said the results could be “problematic” and that more study needed to be done. The post acknowledged that there was concern that some tweets received preferential treatment due to the way users interact with the algorithm by adapting their timeline.

“Algorithmic amplification is problematic if there is preferential treatment depending on how the algorithm is constructed relative to the interactions people have with it. Further root cause analysis is needed to determine what changes, if any, are needed to reduce the negative impacts of our hospitality timeline algorithm, ”the post said.

Twitter has said it will make its research available to outsiders such as academics and is preparing to allow third parties more access to its data, which is likely to put more pressure on Facebook to let him do the same. Facebook is being urged by politicians on both sides of the Atlantic to release its research to third parties after tens of thousands of internal documents, which included revelations that the company knew its Instagram app was harming people’s mental health. adolescents, were disclosed by whistleblower Frances Haugen.

The Twitter study compared the two ways a user can view their timeline: the first uses an algorithm to provide a personalized view of which tweets the user might be interested in based on which accounts they interact with the most and other factors. ; the other is the more traditional timeline in which the user reads the most recent messages in reverse chronological order.

The study compared the two types of timeline by examining whether certain politicians, political parties or news outlets were more magnified than others. The study analyzed millions of tweets from elected officials between April 1 and August 15, 2020 and hundreds of millions of tweets from news agencies, mainly in the United States, during the same period.

Twitter said it would make its research available to third parties, but said privacy concerns prevented it from making the “raw data” available. The post said: “We make aggregated data sets available to third-party researchers who wish to replicate our main findings and validate our methodology, upon request.”

Twitter added that it was preparing to make internal data available to external sources on a regular basis. The company said its machine learning ethics, transparency and accountability team is finalizing the plans in a way that protects user privacy.

“This approach is new and has not been used on this scale, but we are optimistic that it will resolve the trade-offs between privacy and accountability that can hamper algorithmic transparency,” Twitter said. “We are excited about the opportunities this work may open for future collaboration with external researchers seeking to replicate, validate and expand our internal research. “

Leave A Reply

Your email address will not be published.