Lexical scores from an interpretable classifier reveal that moderation criteria differ sharply across subreddits, though the extracted criteria only partially match the accuracy of full-text neural models.
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
-
Decoding the Rule Book: Extracting Hidden Moderation Criteria from Reddit Communities
Lexical scores from an interpretable classifier reveal that moderation criteria differ sharply across subreddits, though the extracted criteria only partially match the accuracy of full-text neural models.