Constrained decoding for seq2seq models improves cross-lingual ABSA by about 5% on the hardest task and over 10% in multi-task setups, surpassing prior methods.
In: 9th International Conference on Learning Representations, ICLR 2021, Virtual Event, Austria, May 3-7
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Improving Generative Cross-lingual Aspect-Based Sentiment Analysis with Constrained Decoding
Constrained decoding for seq2seq models improves cross-lingual ABSA by about 5% on the hardest task and over 10% in multi-task setups, surpassing prior methods.