Home → News → Reuters Built Its Own Algorithmic Prediction Tool... → Full Text

Reuters Built Its Own Algorithmic Prediction Tool to Help It Spot (and Verify) Breaking News on Twitter

By Nieman Journalism Lab (MA)

December 1, 2016

[article image]

Reuters has developed Reuters News Tracer, software that automates the process of spotting breaking news.

The program is designed to monitor Twitter, and identifies what it thinks are emerging reports and clusters relevant tweets into events to generate information and metadata about what the story might be about.

Reuters News Tracer can determine how to identify which events are newsworthy, filter out assertions of opinions from facts, and verify whether the reports are true. Drawing on historical datasets and insights from Reuters reporters, the news organization essentially taught its algorithm to think like a reporter.

Reuters says the tool has paid off in a big way, giving it an eight-minute head start on reporting on the Brussels bombing earlier in the year and a 15-minute jump on sending out news alerts on the Chelsea bombing in New York in October.

The tool also extends Reuters' reach. "We can't be at everything," says Reg Chua, Reuters' executive editor of data and innovation. "Our tool helps shift some of the burden of witnessing and lets journalists do much more of the high value-added work."

From Nieman Journalism Lab (MA)
View Full Article


Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA


No entries found