• Wed. Nov 29th, 2023

News Eyeo

All Important News

Enhanced Algorithm for Detecting “Potentially Inaccurate” Community Notes Gives More Authority to Reviewers


Nov 22, 2023

X (formerly Twitter) has announced a new approach to the algorithm that rates the usefulness and veracity of Community Notes. This will introduce improvements to detect “potentially inaccurate” notes or those that lack support from reliable sources. The changes will be driven by diligent users whose ratings will carry more weight in the algorithm.

Community Notes, also known as Community Notes, are a tool designed to keep users better informed about what they read and share on their profiles. The objective is to add more context to potentially misleading posts. Other users are able to rate these notes to help determine their usefulness. In September, it was announced that a page listing all Community Notes proposals for the same post would be created so users can evaluate its usefulness.

Last October, X reported that Community Notes must include verified sources in order to provide context to users’ publications on the social network. The platform has shared its intention to continue working to ensure that Community Notes provide “broadly useful, clear and accurate” content. The algorithm is being improved to identify notes with accuracy or source support issues.

Improvements have been made to the algorithm’s open source scoring system. This allows it to more accurately detect notes that are “potentially inaccurate” or not supported by reliable sources. The algorithm focuses on notes that have issues with accuracy or source support, identifying them through a wide range of perspectives.

Certain users named “evaluators” will be assigned additional weight in the algorithm. This will be due to their diligent approach in examining and verifying details, allowing for faster and more precise identification of notes with inaccuracy problems. Users who regularly evaluate community notes will be identified and those who are systematic and precise will be given additional weighting in the algorithm. Conversely, those who are too quick in their evaluations or miss errors, will have their weighting reduced.

By Editor

Leave a Reply