▼ Scroll to Site ▼

Newsletters




Using Machine Learning to Report News at Data Summit 2018


Machine learning is changing the way we interact with things and each other. As artificial intelligence gains steam, what we know about the world is changing.

Reuters is introducing a new tool called News Tracer. It is a capability that applies AI in journalism to find events breaking on Twitter. It assigns them a newsworthiness score so people can focus on the events that are important.

John Duprey, senior architect, Thomson Reuters Labs' Center for AI & Cognitive Computing, Thomson Reuters, presented his session, “Reuters News Tracer: The Journalist Bot in Training” on Tuesday, May 22, 2018 during Data Summit 2018. Duprey explained how machine learning and AI is transforming how news is being reported.

Data Summit 2018 is taking place at the Hyatt Regency Boston, May 22-23, with pre-conference workshops on Monday, May 21. Cognitive Computing Summit will also be co-located at the event.

Reuters News Tracer assists journalists with their assignments by helping discover breaking news via tweets, Duprey explained.

Twitter consistently leads traditional media in reporting newsworthy events. However, Twitter is overwhelming and because of the constant content, it is messy and can be unreliable.

AI can address these problems by keeping complex models sparse, it creates a faster and more manageable tool that locates news.

To address the overwhelming aspect, the tool can use boosted filters to identify what isn’t newsworthy. Several filters include language types, profanity, ad/spam, and “chitchat.”

The tool uses real-time clustering to focus on events rather than real-time tweeting. The algorithm can group tweets and manages the information by asking who, what, where, and when.

To determine veracity of the news, News Tracer uses the list feature to mine information to build up a knowledge base of users.

Two data processing levels are involved to verify tweets and keep abreast of the news. Journalists moderate news with an application that reduces clusters of newsworthy information.

“We think the clever and judicious use of AI is how the tool is so successful,” Duprey said.

Data Summit 2019, presented by DBTA and Big Data Quarterly, is tentatively scheduled for May 21-22, 2019, at the Hyatt Regency Boston with pre-conference workshops on May 20.

Many presentations from Data Summit 2018 have been made available for review at http://www.dbta.com/DataSummit/2018/Presentations.aspx.


Sponsors