A GAN-LLM data augmentation pipeline creates the SinaSarc dataset with user behavior and feeds it to an extended BERT model that reaches F1 scores of 0.9138 and 0.9151 on Chinese sarcasm detection, beating prior methods.
Humans require context to infer ironic intent (so computers probably do, too),
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A GAN and LLM-Driven Data Augmentation Framework for Dynamic Linguistic Pattern Modeling in Chinese Sarcasm Detection
A GAN-LLM data augmentation pipeline creates the SinaSarc dataset with user behavior and feeds it to an extended BERT model that reaches F1 scores of 0.9138 and 0.9151 on Chinese sarcasm detection, beating prior methods.