Facebook icon on the phone screen. Photo: AFP/VNA
According to The Verge, in October 2021, Facebook engineers discovered that some content was flagged as suspicious by members of a third-party authenticity checking program but was still selected by Facebook’s algorithm. selected for publication in the user’s News Feed.
The error was so severe that a team of engineers had to create an internal report describing it as a serious failure to rate content. Because the root cause could not be found, engineers monitored this situation until the content rating issue was fixed on March 11th.
Reacting to the information posted on The Verge, Facebook said that the news site technology This article exaggerated the aforementioned content rating error, and asserted that it had no lasting effect on postings identified as “problematic”.
According to a spokesman for Meta (Facebook’s parent company) Joe Osborne, the bug affects only a very small number of views of content posted on Facebook, because the majority of posts in a user’s News Feed are not eligible for downgrade priority level. In addition, other mechanisms set up to limit views of negative content remain, including downgrading, authenticity labeling, and removal of infringing content.
Since December 2016, Facebook has partnered with about 80 organizations to launch the Content Authenticity Check Program on the platform, WhatsApp and Instagram. Content rated “false information” will be demoted in News Feed, so fewer people will see it. If a user tries to share this content, they will be linked to an article explaining why the content is misleading.
If they continue to choose to share content, they will receive a notification with the article link. However, content will not be removed from Facebook’s feed. The fact-checking party has the right to choose how and what they want to investigate.
at Blogtuan.info – Source: Soha.vn – Read the original article here