NoBS - Review Checker

Inspiration

One of our team members was buying a bluetooth speaker, and was reading reviews when he realized that many of them seemed like they came from Chat GPT. That moment, NoBS Reviews was born.

What it does

Our proprietary Neural Network (Almost 90% accurate) will identify which reviews are fake. Then, our extension will tell the user:

1. What % of reviews are fake
2. A letter grade representing the quality of reviews
3. Adjusted rating - the stars on the front of the card represent an adjusted star rating taking only the real reviews into account
4. The back of the card will feature an alternative product with more trusted reviews ## How we built it First, we had to train the neural network on a significant amount of clean data. Once that was done, we spent many many hours figuring out and implementing the algorithm and feature set. Finally, we fixed the front end and the alternate product feature. ## Challenges we ran into So many issues, from github causing problems to the front end not working. ## Accomplishments that we're proud of Neural Network Accuracy ## What we learned So much. The power of energy drinks, mostly. ## What's next for NoBS - Review Checker Adding features based on user feedback :)

Built With

• ai
• html
• javascript
• python