In social media, we daily encounter many cases of mass harassment on platforms like WhatsApp, Twitter, Instagram, etc. It makes a bad social atmosphere.
What it does
We have created a tool to counter the above-mentioned problem. It analyses the messages and logs from different social media platforms and predicts the result about the message or log.
How we built it
We use the ML algorithms like Random Forest Classifier and Count Vectorizer for natural language preprocessing part. For the backend, we used flask microframework for integrating the web part with ML model. For Frontend, we have used HTML, CSS and API from Unsplash for images.
Challenges we ran into
We had a hard time in integrating the backend with the ML model. We also have faced bugs in the model itself. Its also very difficult to work from different places and integrating afterwards.
Accomplishments that we're proud of
We feel in showcasing our tool and the problem that it solves. We also felt the vibes of team in such online event.
What we learned
We learned how to implement our idea in such short time as this is our first hackathon. We learned building and integrating others works as well.
What's next for Combating CyberBullying
We want to not stop it here and take it further. We will definitely make other feature and improve the existing in terms of larger dataset and accuracy.