Tell Me More

Aims

Can we provide more in-article contextual information so that complex news stories are more accessible to a more diverse set of readers?

Outline

We exclude and even shame sections of our audience​ by not including enough context within complex news stories.

Tell Me More stemmed from an ideation session looking at how to contextualise complex news items, and was one of the two ideas taken forward to a full project cycle. The concept was a small expandable explainer block at the top of a news article explaining some of the key concepts that may not be fully explained in the article or where some prior knowledge is assumed. The idea was to give some optional background information for audience members who may not be up-to-date with ongoing complex news stories.

After this initial ideation session, we started work building a prototype that uses automation and machine learning to identify complex terms from articles and generates explanations for them.

Those explanations can then be subbed and used by the journalists to add more contextual content to an article.

The aim of this prototype is two-fold:

1. to provide more in-article contextual information so that complex news stories are more accessible to a more diverse set of readers
2. to provide semi-automated workflows for journalists to create these explainers in order to increase efficiency, consistency and accuracy explanations

Our prototype automatically identifies topics within an article that might require an explanation, and then uses GPT-3 (a natural language AI model) and pre-published BBC News content to automatically generate suggested explainable copy for those topics.

Next steps

We are now turning our focus onto how we can get these explanations in front of audience members for validation.

It has been proposed that we build and test our Tell Me More prototype within the context of the ongoing Cost of Living crisis. It is a current complex topic, and is particularly relevant to our target audience.

Another area that has been discussed is within elections, potentially for the local elections in May 2023.

Results

• User testing the original prototype proved to be very positive
• Interest in using explanations for elections and cost of living crisis