Twitter tries to better predict the stories you want to see

Information discovery is the name of Twitter’s game, so Tuesday the home for short-and-sweet bursts of news announced enhanced tools to better predict more of the stories members want to see.

Twitter has tweaked its Discover tab to feature more “personalized and meaningful” stories. The new and improved Discover area, gradually being rolled out on web, iPhone, and Android over the next few weeks, also has a reworked look that throws additional social context into the mix.

The Discover tab was first introduced in December of last year when Twitter went live with its “new-new” look and feel. At the time, the tab was pitched as a personalized hub for discovering relevant stories of interest. Now, the tab is said to factor in new signals such as content shared by friends-of-friends, and promises to present stories that are even more personalize to your tastes.

“We’ve improved our personalization algorithms to incorporate several new signals including the accounts you follow and whom they follow,” Twitter director of engineering Ori Allon said. “All of this social data is used to understand your interests and display stories that are relevant to you in real-time.”

Twitter first works to identify and rank connections based on importance, then it seeks out URLs shared by that circle of people, URLs are converted into stories, and stories are ranked by tweets and relationships, Allon explained in a blog post on the technology behind the update.

The stories themselves also look a little different. The boxed blurbs now display who tweeted about the stories in question, and members can click to “view tweets” for additional context.

The Discover tab updates will likely go unnoticed by most users as they’re more algorithmic than cosmetic. But if Twitter can use this tab to do a better job at presenting stories people want to see, the improvements could go along way to hooking new users and getting passive Twitterers to engage more on the information network.

Photo credit: ilse/Flickr