Twitter Algorithm Definition

How Twitter's Computers Read Tweets

Algorithm on dry erase board
Algorithm. Jeffrey Coolidge / Getty Images

The smart folks at Twitter have put together a super sophisticated algorithm so that their smart computers know how to "read" the gazillion tweets they're pushing through the firehose.

An algorithm, any algorithm, is used to process data, often putting them into buckets that will give a final conclusion. For example, when you search for something in Google or Bing, the search results that are returned to you come from an algorithm. The search engine's algorithm determined that what you want, based on your search, is what they're revealing to you.

Twitter's algorithm, which reads and sorts Tweets, allows Twitter to figure out what's going on anywhere.

For instance, when Michael Jackson died, his death marked first one, then two, then four, then six of the ten trending topics that afternoon. And, that was twenty minutes before the local radio news station reported it.

At the end of every year, Twitter publishes a Golden Tweets page that displays the most retweeted Tweets of the year. That's data they'd never be able to parse without a specific algorithm meant for counting those retweets.

When Twitter first revealed the new Discover tab, they wrote about the algorithm used to build it:

"We are beginning to roll out a new version of the Discover tab that is even more personalized for you. We’ve improved our personalization algorithms to incorporate several new signals including the accounts you follow and whom they follow. All of this social data is used to understand your interests and display stories that are relevant to you in real-time.

Behind the scenes, the new Discover tab is powered by Earlybird, Twitter's real-time search technology. When a user tweets, that Tweet is indexed and becomes searchable in seconds. Every Tweet with a link also goes through some additional processing: we extract and expand any URLs available in Tweets, and then fetch the contents of those URLs via SpiderDuck, our real-time URL fetcher.

To generate the stories that are based on your social graph and that we believe are most interesting to you, we first use Cassovary, our graph processing library, to identify your connections and rank them according to how strong and important those connections are to you.

Once we have that network, we use Twitter's flexible search engine to find URLs that have been shared by that circle of people. Those links are converted into stories that we’ll display, alongside other stories, in the Discover tab. Before displaying them, a final ranking pass re-ranks stories according to how many people have tweeted about them and how important those people are in relation to you. All of this happens in near-real time, which means breaking and relevant stories appear in the new Discover tab almost as soon as people start talking about them."

Most businesses with algorithms process tons of data every day. The algorithms get updated frequently as needed. For example, in the past several years, Google has updated their search algorithm (to the dismay of SEO's everywhere) a ton of times. What you get as a search result today for any given search is highly unlikely to what you would have found years ago.

Twitter's own search algorithms have gotten incredibly dynamic as well. You can find people who are specifically asking a question, who use a smiley face in their Tweet, and people who are Tweeting specifically in your geographic location.

Twitter's algorithm doesn't need to be as aggressive as Google's, but it's certainly robust and is effectively being used to create new ways to view Twitter's existing data.

Common Misspellings:

twitter algorythm​