In the high-octane media environment that is social media today, it’s sometimes easy to get overwhelmed.
Enter Dataminr, a web- and app-based tool that takes the massive firehose of information that is Twitter and transforms it into a manageable stream of “signals” – or tips – pointing to potential breaking news.
In a two-part video training, Garrett Santora of Dataminr and Mary Nahorniak of USA TODAY talk about Dataminr’s evolution from a tool for the financial and corporate worlds into one that journalists can use every day – and often, all day long.
In the first video, Santora – Dataminr’s news partnerships manager – details how the firm’s algorithms assess every public tweet in real-time, looking for patterns and comparing current tweets to the evolution of tweets from previous news stories. That helps the Dataminr program determine whether Twitter activity around a certain event warrants an alert to its users.
So, for example, the collapse of a construction crane in New York City prompted several tweets from witnesses – and eventually an alert to Dataminr users who had indicated they wanted to see breaking news from New York.
Dataminr users are able to set their own parameters for what they want to see, by region and topic.
“Everyone in the newsroom is using this – but everybody in the newsroom does not have the same focus,” Santora said.
At USA TODAY, Nahorniak is the deputy managing editor of digital and works throughout the day to manage social media and breaking news. Dataminr is constantly on, integrated into her TweetDeck.
In the second video, Nahorniak details how Dataminr was used, from start to finish, in a set of stories. One was international: the 2016 terrorist bombing in the Brussels airport. The second was local: a courthouse shooting in Michigan. In both, Dataminr’s alerts allowed editors and reporters from throughout the Gannett/USA TODAY network to stay on top of a fast-moving story. That, in turn, allowed USA TODAY to alert its readers as the story progressed.