Blogzart-v2.0

The new update introduces a new AI based Comment Flagging System

LSTM-RNN model with 99% accuracy

Now when you post comments they will be first passed through the model for inference and scanning to flag harmful and violent comments. The website using REST Api sends a get request to the model hosted in GCP(Google Cloud Platform). If the comments fall under any of these categories('toxic', 'severe_toxic', 'obscene', 'threat', 'insult', 'identity_hate'), they will flagged and wont be posted. If it doesn't violate any of the guidelines, it will be added with the rest.

Overview

comment that violates the guidelines

user shown warning with flash messages to follow the guidelines and comment is not posted