Machine learning can help newsrooms find stories, advance their goals


Machine learning can help journalists enhance their workflow and find stories that may have been missed otherwise, said Google News Lab teaching fellow Michael Grant during a brunch workshop at the International Symposium of Online Journalism (ISOJ). 

Michael Grant
Michael Grant

The workshop, “Artificial Intelligence and Machine Learning in the Newsroom,” on July 21 was part of ISOJ’s first-ever online-only conference. During the workshop, Grant showed attendees how different tools can be used to gather news, find information, generate story ideas, identify trends, track events and categorize data.  

“You can really start to figure out how to tap into some of the products that are available on the market that leverage the power of machine learning,” Grant said. 

Newsrooms across the world are figuring out how machine learning could work best to advance their teams’ goals. 

An example he gave was that some newsrooms, like Reuters, are using a machine learning tool to search through tweets in order to detect breaking news scenarios. This saves journalists the time and energy of having to search through social media. Instead, the reporters can focus on what they do best––interviewing, investigating, researching, and writing. 

“This is really great for supplemental, where we might not want to throw our resources into things that…[have] a ton of content and that we’d have a hard time even covering with manpower,” Grant said. “This is a great instance where we can throw technology at that particular problem and start to really surface things that might be useful to an audience.” 

In addition, other news organizations, like Vice, are using machine learning in an attempt to make their information more accessible. Vice uses a Google translating system to make some of their stories available in several different languages. 

Meanwhile, ProPublica has used machine learning to track hate crimes in the United States in order to create a central tracking system that didn’t exist beforehand. 

Grant said there are many ways that newsrooms can harness machine learning to make their reporting more efficient, thorough, and unique. Using the technology, he added, can also save reporters time and allow them to focus their skills and resources on facets of the job that technology is not able to do. 

Google News Lab sponsored the conversation about machine learning in English and Spanish. To learn more about tools already available visit Google News Lab’s training tools. Watch the full workshop on the ISOJ YouTube channel, and read more about the workshops on our website.