Table 9 Comparative results between the toxicity classifiers MS Azure Content Moderation API and Detoxify.

From: A dataset of synthetic art dialogues with ChatGPT

Metric

MS Azure Content Moderator

Detoxify

Expert

User

Expert

User

Accuracy

0.522

0.626

0.500

0.553

Precision

0.328

0.363

0.031

0.110

Recall

0.534

0.766

0.492

0.963

F1-Score

0.406

0.493

0.059

0.198

Back to article page