Read more from Technology in Policy.

By: Akash Chandrayan (Faculty of Law, Jamie Millia Islamia), Keira Wong (International School of Bellevue), Krish Shah (Department of Software Engineering, University of Waterloo), Priyanshu Vanani (Redbricks School), and Tamish Shukla (Lotus Valley International School)

Fake news is prevalent in all sorts of ways through newspapers, magazines, and especially the internet. Over time, this has made it difficult to know what can or can’t be believed. The spreading and sharing of misinformation by people through social media also play a huge role in gaining traction with fake news. There are few systems present that are designed to combat fake news and those can be easily manipulated through user comments. To counter this, our team has created a program that will take an article as input and evaluate its authenticity. On top of that, if an article is found out to be fake, the program will suggest better news sources to look into based on your interests.

Introduction

The universe of “fake news” is much larger than simply false news stories. Some stories may have a nugget of truth but lack any contextualizing details. They may not include any verifiable facts or sources. Fake news can be of multiple types:

1. Satire – made-up stories that aren’t meant to be taken seriously; written to joke about the news or celebrities.

2. Clickbait – eye-catching but misleading headlines; designed to get people to click on links to make money or views for a website.

3. Propaganda – false facts which are written to promote a political agenda or a set of ideas.

4. Mistakes – sometimes accidents can happen, but a trusted source will always correct errors in their stories and say when they’ve got things wrong.

There are not currently adequate tools present to help thwart fake news. There are mostly only manual methods of recognizing fake news. The ones that are automated only check some specific articles (e.g. FactCheck.org) and have limitations. A team at Penn State’s College of Information Sciences and Technology showed how these fake news detectors on social media sites can be manipulated through user comments.

These sites such as Facebook are the main target for the widespread use of fake news. In 2014, there was an alarming story that bounced around accruing hundreds of thousands of likes: the small town of Purdon, Texas, was reported as quarantined after a family of five was diagnosed with Ebola. The story was a total hoax, put out by a deeply cynical site called the National Report.

A rather limited number of people are educated on this crucial topic. The media should be an informative haven rather than a medium for spreading fake news. The spread of anti-vaccination misinformation on social media, (and its implications for public health and the global fight against COVID-19) is a textbook example of how misinformation can have serious real-world effects, particularly while we tackle the virus.

ReNews

For people wanting to educate themselves on current events, politics, or any story, the truth is paramount. No one wants misinformation when they read articles, documents, or the newspaper. Our product, ReNews, brings the light of truth into our digital world of information.

ReNews is a product solely built for the purpose to flag websites that contain false information. It recognizes fake news instantly. Using this tool, fake news will be much easier to spot in our day-to-day lives. ReNews’ competitors only contain information on particular articles on the internet, while ReNews is able to dictate the truth on any article it sees.

We have built ReNews in two ways, to allow the user to choose what is best for them. First, we made a website that takes in articles and spits out the truth. For more frequent users that are in a rush, we made a Chrome extension that sits happily in their browser.

Objectives

1. ReNews recognizes fake news almost immediately and indicates the user, as well as uses machine learning to function. This will help improve the system over time. The program searches for flagged websites and articles from a vast database that assists in generating the information needed.

2. ReNews has an added feature that provides reliable and trustworthy sites based on your interests, related to the topic you were searching for. If the original site is found out to be fake, this function will provide aid and allow the user to obtain the result desired.

3. On top of that, once a fake article or site is found, the system could send a detailed report to the concerned authorities immediately. The authorities will take the action that they deem is necessary. Along with that, the database will also be updated from time to time.

Technical Information

To combat the devastating effects of fake news, we utilized machine learning (ML) — specifically linear and logistic regression — to determine the validity of an article. We first constructed the ML model with Python in the Flask framework and deployed it on Heroku. We created a front-end website with ReactJS in which the user could input their article. Later we realized that it would benefit the user experience to have a Chrome extension instead, so we built that using the classic HTML, CSS, and Javascript.

Potential Limitations

1. The primary difficulty is that the user will have to select and paste the text every single time they would like to check an article or a website. In the future, we hope the system would develop to automatically scrape the necessary information off of a website.

2. The second issue that arises is that this is a newly designed model. Hence, there are still chances of falsely labeling a fake news source as real or vice versa. This could be a threat to the validity that our innovation promises. However, we are still improving our model to ensure that this possibility is lessened. As of now, our model has a 96.25% accuracy rate.

Conclusion

The effects of fake news can be catastrophic. Elections are meddled with, businesses fail and individuals are shamed based on a lie.  People who spread fake news do it for their own profit. Strict regulations and actions on fake news should be promptly taken as it is of great concern. However, there are very few of these regulations being implemented. This is where ReNews plays a major role. Our team hopes the system that we have created would enact a substantial change in the fight against fake news. 

With ReNews, we are renewing the face of storytelling. No more media manipulation or suspicious stories — we simply want to bring humanity to real news. The world has too many stories to pollute with fictitious fables. Let’s flip the switch, for real this time. 

References available in full pdf.