A comparative annotation study finds ChatGPT over-identifies inappropriate targeting in Reddit conversations and uncovers four new target categories beyond the standard hate speech classes.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Understanding and Analyzing Inappropriately Targeting Language in Online Discourse: A Comparative Annotation Study
A comparative annotation study finds ChatGPT over-identifies inappropriate targeting in Reddit conversations and uncovers four new target categories beyond the standard hate speech classes.