Why integrating image verification in the newsroom matters
Images always increased the power of a story. And this never more so than today. The proliferation of smartphones and digital cameras means that when news occurs, the first person to capture it will likely not be a journalist. User-generated content is flooding newsrooms, however, it seems newsrooms struggle with verification of this content, frequently misunderstand it, or — at worst — ignore it completely.
The immediacy of reporting, however, means that journalists are ever more pressed to publish powerful images at speed - before their competitors. But they also need to know they are right. To verify them. There are many tools out there to support image verification, but often journalists don’t know them or, at the very least, don’t know how use them to their full capacity.
How do we make sure images sourced by journalists are correctly verified and put in the right context? How can we make it a consistent process that allows newsrooms to incorporate citizen photography with less risks?
Hence the Verified Pixel Project, a service that aims to help demystify image verification for journalists.
Creating Verified Pixel
The Verified Pixel Project is a partnership between John S. Knight Journalism Fellow Samaruddin Stewart , Eyewitness Media Hub co-founder Sam Dubberley, and Douglas Arellanes, co-founder and director of innovation at Sourcefabric.
The core team is supported in the implementation of the project by Nuno Vargas and Mark Lewis.
The project was recently awarded a Knight Foundation Prototype Fund Grant. In the next few months, the project team plans to launch Verified Pixel as a prototype, but they also intend to reach out to newsrooms for beta testing and further development. The Verified Pixel Project will build on Sourcefabric Superdesk technology and has been conceived with the focus of being simple to use, responsive, and easy to integrate into the workflow of newsrooms.
“Verified Pixel is interesting for Sourcefabric because it leverages a lot of the work we have already done in citizen journalism verification with Citizen Desk, as well as a lot of the deeper technical work we’ve built into Superdesk. We’ve said that Superdesk is the new open source code base for journalism, and Verified Pixel is a good way to show just that.” says Douglas Arellanes.
“We feel a key benefit of the Verified Pixel Project is creating an automated workflow to support image verification. Automation of verification tasks will help to assess which photographs can be most easily verified from photographs which can be dismissed as misrepresented or manipulated quickly - think of it as ‘image triage.’ This will allow newsrooms the ability to refocus energy and time on steps that are not (yet) automated, such as contacting the uploader or visually identifying landmarks,” says Samaruddin Stewart.
“The pressure to publish means that many of the simple steps required for verifying an image get either forgotten or skipped. One of the goals of Verified Pixel is to ensure that these steps are run every time, consistently. We feel this will be a real benefit to the journalist on deadline pressure,” says Sam Dubberley.
In an increasingly fast-paced digital news environment, sharp verification and fact-checking skills on user-generated content have become essential. In order to really work, the process of image verification needs to be more accessible to non tech-savvy journalists and more consistently automated in the newsroom workflow.
How does it work?
With a single interface, journalists will be able to consult different tools via layered API calls. That allows the journalist to answer simple verification questions such as: can I geo-locate the image? do weather conditions on the image correspond with reports on that day? has the image been modified with image alteration software? what information do I have about the source that sent me the image?
It is clear though, that Verified Pixel Project will not substitute the work of social media editors. At the end of the verification process there should always be a trained individual who is able to exercise editorial judgement on the information available. Verification of images can only happen in the context of newsrooms already equipped for integration of user-generated content.
Stewart added, “I’m honoured that the Knight Foundation is providing a runway to make a prototype tool to potentially help newsrooms solve this issue and scale to the growing demand. Looking forward, this tool could easily impact other industries that value visual verification, including human rights/humanitarian agencies, governments, and even the general public.”
The Verified Pixel prototype will be presented in mid-October to the Knight Foundation, but if you want to follow the developments around the project, keep an eye on www.verified-pixel.com or get in touch with Sourcefabric!