以下内容由机器翻译生成。如果您觉得可读性不好, 请阅读原文或 点击这里.
After facing a barrage of criticism over the toxicity of its community, Twitter announced last month that it will attempt to more actively measure and combat bad behavior, with input from outside researchers. We’re now seeing an early, straightforward idea: make its rules of conduct more visible.
CEO Jack Dorsey tweeted that the company has partnered with an outside team, led by researchers Susan Benesch and J. Nathan Matias, to run a test, starting today. “Research has shown that when institutions publish rules clearly, people are more likely to follow them,” the researchers wrote in an announcement. With that in mind, the researchers will test whether making users more aware of the rules will promote civil discussion on Twitter.
Launching a new study today with @SusanBenesch 和 @dangerousspeech to help diminish abuse on Twitter. It’s a simple idea, and we have an open and accountable way to share the findings. https://t.co/aH2KCWJiRd
— jack (@jack) April 6, 2018
The exact mechanism of the test isn’t clear. The researchers write that they “cannot describe it in detail now without jeopardizing the integrity of the results,” but say they received approval from two university ethics committees, and filed the outline of the plan with a neutral third party. They will work only with anonymized data.
Twitter announced its plans to improve its platform after a deluge of criticism, as users complained about issues like harassment and misinformation. While, in recent weeks, the focus of scorn has shifted to Facebook, Twitter hasn’t escaped entirely: the company posted a blog post defending its moderation process after hoaxers attempted to spread lies on Twitter about this week’s shooting at YouTube headquarters.
The researchers say the ultimate results of the new test will be published in an academic journal. “We hope that this project will supply practical knowledge about preventing abuse online, and that our process will inspire further transparent, independent evaluations of many other ideas for reducing online abuse,” they write.