A web-based encrypted group chat system which has the functionality of detecting instances of cyber bullying/offensive messages. The chat system uses asymmetric and symmetric cryptography to encrypt the data. A group key is used to encrypt communications and the group key is shared amongst the users using asymmetric cryptography.
The problem statement is to design a model which can effectively detect bullies and stop them while securing the privacy of the users which is important as well.
- Clone the repository
- Install the python dependencies by executing
pip install -r requirements.txt - Execute the code cells in the python notebook NN_detector.ipynb which generate the model weights and saves them in Bullying Model directory
- Install the NodeJS dependencies by executing
npm install - Run the server by executing
npm run start
-
Neural Network Layers
-
Neural Network Accuracy
-
Chat system demonstation
