The Perspective API giving toxicity (and other models) scores for Hamlet (Act III, Scene I)
#datascience #dataviz #pushshift
Conversation
Perspective finds that rather incoherent.
1
Replying to
I suppose they are not. I can imagine a simple method to estimate a measure of uncertainty on the scores.
It would definitely help comparing scores. Without uncertainty, I’m not sure how to compare scores among samples.