Related Fact Checks

Sreya Guha

Related Fact Checks was a service that provided fact checks related to an article. If a user was reading an article and not sure about the claims the article is making, Related Fact Checks (RFC) could give them relevant fact checks. RFC did not automate fact checking; instead, it provided a list of relevant fact checks available to allow the user to make an informed decision. RFC is no longer available as a service but you can learn more about the motivation and design on this page. For a demo and explanation, you can watch this video. To learn more about how this service was built, you can read this paper, which was presented at the International Semantic Web Conference in Vienna in 2017.

Context

Digital disinformation, more colloquially known as fake news, rose to prominence in the 2016 election. Though the intent to mislead the public for political and social gain has always persisted in society, the rise of the internet and subsequent innovations from the information age have allowed for the undermining of democratic processes.

While fact checks emerged as a traditional countermeasure to fake news, they still required the user to seek out such information. There is a missing gap between fake news and fact checks, minimizing the possible effect of the latter.

I built RFC to attempt to bridge this gap. I wanted the tool I built to support critical thinking rather than replace it. I built RFC to give users more context and encourage this deliberation. Given a link to a website for a story, RFC provides the user with a related fact check from a corpus of fact checks (collected by using the Schema.org structured data markup on fact check pages). For stories without fact checks directly addressing the claim, RFC aimed to show stories with the same theme to provide the user with more context.

While the vast majority of misleading content has not been fact checked due to the relative ease of creation and cost of fact checking, my research throughout this project revealed that stories follow certain patterns. In particular, analysis showed that there are a small number of themes (e.g., anti-vaccine, anti-climate, etc.) that repeatedly appear in many misleading stories.

Here is an example of RFC at work for an article claiming that the American College of Pediatricians linked the HPV vaccine to serious health conditions. RFC provides the user with 5 relevant fact checks and as shown, these fact checks address the claim.

Press and Awards

RFC was featured in this article in Poynter in 2017. RFC also received the 2017-2018 Cutler-Bell Prize in High School Computing awarded by ACM (Association for Computing Machinery) and the Computer Science Teachers Association (CSTA). The press release can be found here.

About me

My name is Sreya Guha and I am a senior at Stanford University, majoring in Symbolic Systems with a concentration in Artificial Intelligence. My Google Scholar Profile can be found here.

More Resources to Combat Misinformation

Many organizations and journalists are working to combat misinformation. Here are a few resources to learn more:
Duke Reporters Lab
International Fact-Checking Network at the Poynter Institute
The Trust Project Center for Media Literacy
Social Media and Fake News in the 2016 Election
Welcome to the Era of Fake News
The Implied Truth Effect: Attaching Warnings to a Subset of Fake News Headlines Increases Perceived Accuracy of Headlines Without Warnings
Evaluating Information: The Cornerstone of Civic Reasoning