Twitter to improve relevancy of feed with machine learning

Twitter is looking to leverage machine learning technology to push high-quality posts to users in their feed, according to CNBC. It will also roll out a feature for users to follow an interest.  The feature aims to filter out irrelevant content and show users relevant tweets from various accounts based on the interests they follow. The feed will be curated by Twitter’s algorithms.Twitter confirmed the news to Marketing.Speaking at the company's US headquarters, a Twitter executive said that the feature is currently being tested with some app users on Android and will be made available to the rest of the users by the end of the year.In a conversation with Marketing in June, Sarah Personette, vice president of Twitter global client solutions, said Twitter is evolving itself to make the platform for conversations. Besides rolling out a feature for users to subscribe to conversations, it has also been testing coloured labels to threads to make them easier to follow and a “ice-breaker” feature where areas of interests are pinned at the top of profiles to encourage new conversations. Twitter, Personette added, has been at the heart of conversations around “cultural moments”, ranging from pop culture to cultural celebrations.Last month, Twitter also unveiled a revamped Twitter.com with new features such as a fresh look, faster navigation and personalisation options. The changes, according to Twitter in a separate blog post, “make conversations easier to find and follow and just a bit more fun”. They include bringing content found in Twitter apps to the website, so users can expect more live video and local moments personalised for them. Additionally, they can get context with profile information within conversations and check Top Trends in any view. Direct messages have also been expanded to allow users to see their conversations and send messages all from the same view.