Harrassment-Detector

This is an attempt to strictly moderate online speech and fight harassment. After testing across multiple servers and with peers and testers of varied views, we concluded that most of the obfuscated speech has been flagged, better than most online models - including Instagram, TikTok, and Discord.

Steps To Use:

1. Git clone this repository by running git clone https://github.com/preethamcoder/Harassment-Detector.git
2. Change into the directory by running cd Harassment-Detector
3. Now, you will need to install the requirements by running pip3 install -r requirements.txt
4. To obtain the model and datasets we used, you will need to request access at preetham23415@gmail.com. We shall get back to you based on the order of emails received.
5. Once you have the model downloaded, simply run python3 app.py and open your localhost to experience this.

Requests From Developers:

Please do not misuse this, make sure to credit the creators of this if used anywhere.

Acknowledgements:

We would like to sincerely thank our friends for allowing us to use their server logs and chats to build our datasets. A huge thanks to Dr. Lipeng Wan, our Cloud Computing professor at Georgia State University, for providing us with the opportunity to design this project. A special mention to Dr. Yi Ding, my personal mentor in this area of work.