active

The Explainer Builder

An experiment in using a single source of truth for journalists creating explainers.

Aims

Can we add to audience understanding of major stories while making it easier for journalists to regularly explain background facts?

Following an earlier News Labs explainers project, we wanted to take the next step by building a fully-functioning newsroom prototype to help journalists avoid having to explain the same concepts again and again.

The system allows newsroom teams to build and maintain a centralised question and answer repository.

Text input matching to questions.

The newsroom prototype tool allows journalists to easily update the answers within explainers.

By having a prototype tool that can repeatedly produce tailored embeds based on the questions and answers on a topic, there is an incentive for journalists to keep the content up-to-date.

Journalists can curate the questions that are displayed in the embed, along with a search box that allows readers to ask questions.

Could this improve audience understanding and engagement?

Our thinking is that including all of these questions in stories will allow our audiences to easily find background information on long-running stories which they may not be familiar with.

Initial analysis is promising and shows that readers are mostly clicking the questions that are chosen by the journalist in the embed.

Work is ongoing to capture the input of users and feed this back to journalists creating the questions and answers. In this way a complete conversational journalism loop could be created in this project.

Can we answer your question?

The prototype features free-text matching, allowing users to input questions and hopefully get relevant answers. All without leaving the story.

Text input matching to questions.

An example of how user input is matched to relevant questions.

Example explainers

  • Coronavirus: available in this story.
  • US primaries and election: available in this story.
  • Climate change: available in this story.

Next steps

We are looking at continuing this experiment with support for multiple languages and perhaps multiple formats - all with this one central repository of questions and answers acting as a single source of truth.

Tweet us @BBC_News_Labs if you want to find out more.

Results

  • Initial results were positive. Audiences engaged with embeds featured on stories published on the BBC News website in February 2020.

Careers

Love data and code?

We're looking for talented developers to join our team.