-
-
Detect Fake news with funny gifs
Inspiration
Most of these days there is a lot of fake news that is circulating through web. If we can make a spam detection for our email then we can also try to make a fake news detecting system. if we have such system, it can raise a flag when you read something which is unreliable.
What it does
First of all, you past the news article and the model will analyze through it and predicts whether the news you should trust or not.
How we built it
The Frontend of the program uses Flask for web deployment and tensor flow and keras for the machine learning. I also added api call to Giffy which provides funny gifs based on the prediction.
Challenges we ran into
The main challenge I ran into is the data. then to process it. i did not know about the tokenizers and NLP to process text to numbers. Even though the accuracy of the model is 99% the validation accuracy is 65% which is very bad.
Accomplishments that we're proud of
I learned about tokenizers which is used in NLP. The meme gif generator based on the prediction.
What we learned
I learned about Recurrent Neural Network(RNN), NLP and Flask.
What's next for FakeNewsDetector
the data processing can be done more precisely and removing the bias the model can make. Deploying it to Heroku and might be its own domain
Log in or sign up for Devpost to join the conversation.