Explained: How Facebook's 'Trending Topics' Work
Explained: How Facebook's 'Trending Topics' Work
Following the controversy about politicial bias in 'Trending Topics' Facebook has made public its internal processes.

Facebook has emphasised that it does not permit its employees to block news stories from its 'Trending Topics' list based on political bias, amid a controversy over how the social media superpower selects what news it displays.

Technology news website Gizmodo had reported that a former Facebook employee said workers "routinely suppressed news stories of interest to conservative readers" while "artificially" adding other stories to the trending list. The story triggered a reaction on social media, with several journalists and commentators raising concerns about alleged bias, and prompted a US senate inquiry.

In a post published to the Facebook Newsroom on Thursday, Justin Osofsky, Facebook's vice president for global operations, outlined its "Trending Topics" guidelines at length. Here's how Facebook's 'Trending Topics' work:

- Trending topics were introduced in 2014 and appear in a separate section to the right of the Facebook newsfeed.

- Though Facebook hasn't said how many people are responsible for the trending topics team. A Guardian report said the team was as few as 12 people.

- According to Facebook, potential trending topics are first determined by a software formula, or algorithm, that identifies topics that have spiked in popularity on the site.

- Next, a team of trending topic staffers review potential topics and confirm the topic is tied to a current news event.

- They then write a topic description with information corroborated by at least three of 1,000 news outlets and apply a category label to the topic

- The team also checks to see whether the topic is covered by most or all of ten major media outlets. Stories covered by those outlets gain an importance level that may make them more likely to be seen.

- Each Facebook user's trending topics are then personalised via an algorithm that relies on information about the user such as "Likes" and their location.(With inputs from agencies)

What's your reaction?

Comments

https://shivann.com/assets/images/user-avatar-s.jpg

0 comment

Write the first comment for this!