Twitter is tweaking its search results so that tweets will be ranked on relevance instead of reverse chronological order.
Twitter says that based on early trials, the filtering of search results by relevance, rather than reverse timeline, results in better engagement and time spent using the service.
Twitter senior software engineer Lisa Huang said that users risked losing engagement with various content types by prioritising one result type.
‘Overall, we’ve observed that people who have experienced this new search results page tend to not only engage more with the search results, but also tweet more and spend more time on Twitter’
– LISA HUANG
“This is especially evident in the case of tweet results versus account results,” she said.
Bringing clarity to the noise of Twitter
After various experiments, Twitter discovered that surfacing relevant tweets to the search as opposed to a straightforward chronological order resulted in greater engagement.
Machine learning has also been employed and Twitter can now train machine-learning models to predict how likely a tweet is likely to be engaged with in terms of retweets, likes and replies.
“Using this information, we can train machine learning models that predict how likely a tweet is to be engaged with (retweets, likes and replies). We can then use these models as scoring functions for ranking by treating the probability of engagement as a surrogate for the relevance of tweets.”
Huang said that Twitter is also trying to surface a more diverse set of results.
“Overall, we’ve observed that people who have experienced this new search results page tend to not only engage more with the search results, but also tweet more and spend more time on Twitter,” Huang said.