Created by Rohan Koodli, Vineet Mudupalli, Vikranth Srivatsa, and Aubhro Sengupta
Winner: #HackHarassment
Winner: .Tech Best Hack with .tech domain name
Check out the Devpost link here.
We were inspired by the recent spread of false news and we also wanted to create a hack against cyberbullying.
Incredible Filter gets rid of any false news and alerts you if you try to tweet a false news story. Also, any hateful tweets are hidden from view. This helps normal people distinguish between what is real and what is fake
We used IBM Bluemix and scikit-learn to write our machine learning backend to classify news articles as fake and to classify tweets as hateful. We used Firebase to store tweet info, and we accessed this through the iOS frontend.
It was difficult to do the machine learning on the news articles, getting Firebase to work with Python 3, and linking the iOS frontend to our backend server.
We wrote a machine learning SVM with scikit-learn and Bluemix.
We learned an easier way to use IBM APIs as well as how to use the IBM Tone Classifier.
A stronger machine learning backend.