In January of this year, Twitter launched Birdwatch, an experimental fact-checking platform, as a way to combat misinformation. The project allows users to report tweets (posts) on Twitter as misinformation, providing references to substantiate such an assessment.
Over the past week, researchers at fact-checking organization Poynter analyzed more than 2.600 reader responses on the Twitter platform, and processed 8.200 ratings. The fact-checker from Poynter also used a public algorithm that the company uses to rank posts based on their “usefulness.”
According to this analysis, the results are not very promising, as it was found that the reactions of readers through the Birdwatch platform were based on unreliable references. However, according to this analysis, it turns out that users who reacted through Birdwatch were not based on references at all.
Through this analysis, the Poynter researcher expressed doubt that automated fact-checking, such as the Birdwatch platform, could solve the problem of fake news. The researcher listed a number of suggestions for Twitter for Birdwatch:
For resources:
• Twitter should require Birdwatch users to cite at least one source within their notes. More than half of the content reviewed in the program did not include a single source.
• Rely on human fact-checkers to check Birdwatch records. Wikipedia, another human-sourced portal, was one of the most frequently cited sources in the program. Fact-checkers do not rely on Wikipedia as a reliable source.
• Promote posts that provide fact-checked articles. URLs from professional fact-checking organizations were rare among the posts I analyzed, even though some of the topics were covered extensively by fact-checkers.
About the algorithm:
• Stop promoting Birdwatch notes that carry misinformation or lack sources.
• Review the verification process so that multiple users who have spread misinformation on their timeline are not allowed to become Birdwatch users.
(Article prepared by Alex Mahadevan, senior multimedia reporter at Poynter)