Neural News

• 

  Home Page

• 

  Applied Tag Page

• 

  Login Page

Inspiration

What inspired us to create this was that as computer science students it is crucial to stay up to date with everything that is going on in the tech world due to its fast pace and exponential changes. When you either look up "tech news" or go to a technology news website, articles range from "I spent a day letting AI make decisions for me" to "Google is getting sued". Another contribution to our inspiration was that we wanted to honor how far technology has come by paying tribute to the traditional newspaper style by implementing it in our user interface.

What it does

Our project summarizes tech related news articles into an optimized list giving the reader a clear and quick understanding of what the article is about, while also including the link to the original article to give the user the option to read the entire article. This application also allows the user to choose what they are looking to read about whether its company based or topic based. In order to save and or share the articles, the user must create or already have a Neural News account.

How we built it

We built Neural News using React as our frontend framework and Firebase as our database. For the backend, we used Python to run the Hugging Face "facebook/bart-large-cnn" model for article summarization. The backend is deployed and optimized for Intel Tiber to enhance performance, and the website is hosted on a .tech domain.

Challenges we ran into

On the frontend, we encountered formatting issues when connecting to the backend. Midway through the hackathon, we decided on a complete theme change, requiring us to rebuild the frontend from scratch.

On the backend, we faced several challenges, including setting up a locally hosted machine learning model for article summarization, finding reliable sources that didn’t block web scrapers, and ensuring the backend provided all necessary data to the frontend. Additionally, learning and optimizing for Intel Tiber to boost model speed was a steep learning curve, but ultimately rewarding.

And as if the coding challenges weren’t enough, we had some unexpected "wildlife encounters" when a rat decided to chase us down as we worked late into the night. To top it all off, a few team members came down with food poisoning, adding some “extra spice” to our hackathon experience.

Accomplishments that we're proud of

2/3 people in our group were competing in a hackathon for the very first time with this project, so we are really proud of stepping out of comfort zones and working with each other in a productive manner.

What we learned

We learned that quick and clear communication is crucial when doing a big project due to everyone having different a different idea on how they want the project to look in the end.

What's next for Neural News

Neural News is hoping to have articles from more news outlets rather than just the major outlets as well as having an option for users to see their local tech news as well. Another addition that would be coming to Neural News is different tags such as global, legal, business.

Built With

• node.js
• python
• react
• tiber