A test-time code-switching and alignment-augmentation framework improves cross-lingual aspect sentiment triplet extraction, outperforming ChatGPT and GPT-4 on four low-resource languages.
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
-
Test-Time Code-Switching for Cross-lingual Aspect Sentiment Triplet Extraction
A test-time code-switching and alignment-augmentation framework improves cross-lingual aspect sentiment triplet extraction, outperforming ChatGPT and GPT-4 on four low-resource languages.