The majority of people receive their news through the internet. However, there is a plethora of misinformation and clickbait titles. We wanted to build something that would easily and effectively combat some of these issues while keeping the public informed.

What it does

It’s a chrome extension that provides information about news articles such as a prediction of if it has a clickbait title, a short summary of the text, and relevant keywords.

How I built it

To detect if an article is clickbait, we used a prebuilt NLP machine learning model and connected it to a Chrome extension via a Python Flask web server. The Chrome extension utilizes JavaScript to retrieve article data from the user’s browser and send it to the Python Flask backend. There all our data processing is handled to detect if it’s clickbait and to generate summaries and keywords. Then, the data is sent back to the chrome extension which displays it onto the browser.

Challenges I ran into

The biggest challenge that we ran into was deploying our backend onto a cloud service. We tried using Heroku, then Azure, but still ran into issues in the end. We also struggled to utilize the use the NLP model

Accomplishments that I'm proud of

Developing an application that we could interact via a google chrome extension was new and rewarding. We realized that a chrome extension would be the best solution for what the application provided. In the end, we committed to this idea and it led to a promising product.

What I learned

We attempted a machine learning algorithm that none of us were familiar with prior to the beginning of DubHacks. With some time we managed to utilize the model which was the biggest thing we learned from this project. This was out first time building a chrome extension, so it was exciting to learn how we could apply our JavaScript experience in powerful ways.

What's next for CheckBait

Next, we would like to improve the visual clarity and design of the chrome extension. We would also like to be able to host the backend on a cloud service, so that we can publish the chrome extension to the Chrome Web Store. We would also love to improve the speed with caching and concurrency to improve the usability.

Share this project: