The goal of this project is to investigate into the performance of different classifiers trying to classify sexually harassing Tweets and thereby, create an end to end fully deployed classifier to predict the same with the best possible results.
Dataset has been created by scraping tweets from Twitter and manually annotating them.
Check Models folder in Code to check the Classifiers.
The FastAPI and Streamlit code for the above project is available here https://github.com/sohamtiwari3120/cyberpolice-fastapi-streamlit.
A new web app using FastAPI and front end using HTML, CSS and JS will be coming in this repo soon.