Hatespeech Research - Algorithmic transparency techniques for NLP


I am working on a research project looking at Hatespeech in Facebook groups. We would like to be able to visualise the diversity of opinions in a group. So where our sentiment analysis indicates negative sentiment toward some ethnic group, we would like to visualise whether that was total negative sentiment or whether any members of the group spoke up against the Hatespeech.

Your suggestions would be most welcome :slight_smile:

Many thanks for your time.

© Copyright 2013-2019 Analytics Vidhya