Twitter Using Deep Learning To Recommend Tweets

With Facebook, Instagram, and Twitter, at the start the companies relied on chronological order to display updates on the user’s timeline. However we later saw a change where these social media platforms started to use algorithms to display posts that are based on the user’s preferences instead of chronological order.

Now it seems that Twitter could be taking things to the next level by using deep learning technology to recommend tweets to users, according to a report from CNBC. In a blog post on Twitter, the company’s engineers explained the decision behind the move to using deep learning technology, as opposed to the previous system which relied on algorithms to determine which tweets might be of interest to you.

According to the engineers, “The company is evaluating and scoring thousands of tweets per second to determine what’s worth recommending in timelines, taking into consideration an increasing number of factors, including whether tweets contain images or videos, the number of retweets and likes, and your previous interactions with other account holders.”

Ultimately what they found was that this helped to increase engagement and time spent on the platform, which is something that many social media platforms are aiming for. “More importantly, online experiments have also shown significant increases in metrics such as Tweet engagement, and time spent on the platform.” So if you are starting to find more relevant tweets shown on your timeline, you have deep learning to thank for that.

Tyler Lee
Categories: General
Tags: AI (Artificial Intelligence), Twitter