The WHO says the 2019 coronavirus disease (COVID-19) epidemic and response has been accompanied by a massive ‘infodemic’ - an over-abundance of information – some accurate and some not – that makes it hard for people to find trustworthy sources and reliable guidance when they need it.

What it does

A platform that enables trained citizen fact-checkers to field early warning alerts in social media and reply natively in "coordinated" fashion with fact checks and authoritative content.

In Leticia Bode's talk at WHO's Infodemic Management consultation, she mentioned user-directed corrections. The social media companies are failing with respect to misinformation and the Infodemic, and the solution has to include the crowd. I imagine a swarm of citizen fact-checkers independently correcting false claims. For a given claim, there is no discernible consistency in content or coverage. The better end of the spectrum of fact-checkers would give links to expert pages alongside messaging that meets the errant poster where they are at. The worst of them would either give no link or a unhelpful link along with off-putting messaging. There is no coordination other than perhaps some Facebook groups.

But what if there was a tool called CrowdFact, a human AI collaboration for fact checking that enabled citizen fact-checkers, journalists, media companies, fact checkers, Internet platforms", and **NGOs to coordinate on fact-checking and steering people to trustworthy information on COVID19 (or any other Infodemic topic). The tool is a browser overlay that creates an up-to-date layer of contextual information on all relevant content, generating a relational knowledge map, enabling users to interact on the same web pages in real time. We call this layer the OverWeb. Our full-stack Covid19 Infodemic Management tool detects emerging fake news; alerts media, fact checkers, and citizen fact-checkers; makes bridges that attach trustworthy evidence to misinformation; and displays fact checks and trustworthy information as citizen fact checkers browse the web for misinformation, imposters, bots, fake accounts, scams, outdated recommendations.

How we built it

We built it with React, Laravel, and Postgresql

Challenges we ran into

We had a some issues finding high impact content involving Trump's gaffe about disinfectants on Twitter without AI. Trying to do this without AI would not even begin to scale.

Accomplishments that we're proud of

The concept, application, and the business model. The concept is ground breaking in that creates the possibility of a coordinated fact checking response among people who don't even know each other. This is made possible by the concept of a bridge and the collective knowledge map into which they aggregate. The application to COVID19 and pandemics will save lives. The business model focus on providing a tool to citizen fact checkers means that news consumers do not have to install an overlay to be impacted.

What we learned

We learned about the multiple points where human-AI can work in this digital tool. We see the package of what we offer as the people, the tool, and the collective knowledge map. We saw first hand how timely, high-precision fact-checking per the WHO Infodemic Management consultation (user-directed corrections, authoritative sources) can potentially change behavior and save lives.

We also have abstracted the capability first demonstrated here with CitizenFact as a pattern for monitoring and intervening on the Internet in situations that require timeliness and precision, like dealing with posters of false claims.

What's next for CitizenFact

We're raising funds to build the beta version of the tool. We are looking for grants and sponsorship for the CitizenFact project. We are a 501(c)(3) project in the United States. If you know anyone who may be able to help, please do let us know.

You can use the alpha version now; register at Join the COVID19 InfoIMMUNINTY Project.

Participate Infodemic Design Challenge.

Share this project: