Credit: By pagedooley on Flickr. Some rights reserved.
When people think of data journalism in the newsroom, they generally refer to two things: finding stories in data, and coming up with the best visualisation to tell the story.
Of course, not every story will lend itself to a data approach, but when it does, spreadsheets and number-crunching can be daunting to the uninitiated.
Last year, the Data Lab at the University of Washington developed Voyager, an open source tool that wants to help people explore and analyse their data by providing recommended visualisations.There is a distinct lack of free and open source tools that try to bridge the gap between data visualisations and data analysisIrene Ros, Bocoup
The Bocoup Data Visualization team recently received funding from the Knight Foundation's Prototype Fund to work on Voyager. Irene Ros, director of data visualization at Bocoup, who is working on the project, said the tool has the potential of "filling an important space". "
"There is a distinct lack of free and open source tools that try to bridge the gap between data visualisations and data analysis, both of which are currently harder for people to do."
She explained that data analysis is usually reserved for people who, for example, "have a statistics background and know how to program in Python", whereas with data visualisation, "you generally already know what you're trying to demonstrate".
Voyager wants to combine the two through a recommendation system, monitoring the user's preferences to suggest data visualisations, as well as explain why a suggestion has been made.
"In general, recommendation engines are something that's very opaque to the user. If you get recommended a book on Amazon, sometimes you know why, but other times you're not so sure.
"We want to make sure that the recommendations make sense with what the users' intentions are and if not, they can divert from that path. But the aim is empowering people to understand the data and be able to direct how it gets displayed to them in the future."
A video demo from the University of Washington of the Voyager prototype, which the team want to make more accessible with the Knight Foundation grant
At the moment, Voyager's prototype allows users to add their own data sets or play around with existing sets.
Data can be filtered by including or excluding certain parameters and the tool will display a variety of visualisation options based on the data.
For journalists, even those not that familiar with data, Voyager can be helpful in grouping together and displaying elements in, perhaps, a less obvious way, which could then become a new hook for a story.
Ros pointed out the "big rise towards more data-driven journalism" often requires a journalist and a data analyst to collaborate.Hopefully, it will bring the journalists closer towards working with the data and engineering teams and help with the conversation between either building more interactive visualisations or telling a more data-driven storyIrene Ros, Bocoup
Voyager will aim to "empower the journalist to do that in their own time", she said, by simply uploading a CSV file from a Google spreadsheet into the platform, which will automatically analyse it.
"Hopefully, it will bring the journalists closer towards working with the data and engineering teams and help with the conversation between either building more interactive visualisations or telling a more data-driven story."
She said that the visualisations are not currently embeddable or interactive, although both options could be considered in the future. The goal is "not so much to be these impressive graphs you can include, but more of a way for you to reflect on your data".
Because the team already has a head start by working to expand an existing prototype, Ros said they are hoping to generate interest from the broader data visualisation community and implement people's suggestions as the project expands.
"We hope this tool will be helpful to the journalism community, so we would love to hear from people who are interested in trying it out, being a part of our studies or who have feedback on the tool in its current form."
Update: This article has been updated to reflect that the Data Visualisation team at Bocoup was awarded funding by the Knight Foundation to develop Voyager.
Free daily newsletter
If you like our news and feature articles, you can sign up to receive our free daily (Mon-Fri) email newsletter (mobile friendly).
