In June 2019, BBC News Labs and Connected Studio brought together over 40 professionals from the fields of language technology and journalism for a #newsHACK looking at tools to help multilingual newsrooms.
The teams were not all preformed. Many came together from different backgrounds and organisations, including the Financial Times, Deutsche Welle, The BBC, academic institutions including the University of Edinburgh and a number of startups.
Participants also heard talks from experts about the latest challenges and opportunities in the field. They were also given access to some of the BBC’s internal tools which they could use alongside their own technologies.
The two day event concluded with each team pitching their new ideas, and receiving feedback from a panel of experts including Judy King of BBC Monitoring, Dr Khetam Al Sharou of University College London and Hernando Alvarez of BBC News.
What each team built
A way to adapt articles between languages without a pure word-for-word translation taking into consideration cultural context. The tool uses entity recognition to link to online resources as well as a a photo gallery to inform journalists in other languages who may have less relevant cultural knowledge.
The team wanted to monitor the way that stories spread throughout different languages and different parts of the world. They followed where stories spread, and allowed readers to see stories in different languages. Their tool also aimed to show journalists a map of where a story has travelled to, and data about its impact.
A way for readers to question and correct machine translations in order to make translations more useful and machine translations better. The team wants to encourage readers of multiple languages to help one another understand articles all while building glossaries to improve the automated translations and source articles.
The team advocated imagery as a way to transcend language divides, particularly for languages where news outlets do not have a specific language service. The team suggested a new image-led news site that makes use of iconography and photographs to convey news from around the world.
This team designed a way to help journalists working with machine translations of foreign language articles by tracking corrections and creating topic specific glossaries. This would stop machines making the same mistakes repeatedly to the frustration of journalists in the newsroom.
Are you working on a related project? Interested in learning more? Have ideas you want to talk to us about? We’d love to hear from you!
You can follow us on Twitter for updates on our work or else send us an email at firstname.lastname@example.org.