A pytorch version of the EMNLP 2022 Paper "A Span-level Bidirectional Network for Aspect Sentiment Triplet Extraction" https://arxiv.org/abs/2204.12674
-
This code corresponds to the manuscript article in emnlp2022 and differs only slightly from the article "Span-level Bidirectional Cross-attention Framework for Aspect Sentiment Triplet Extraction" on Arxiv.
-
This code has many experimental settings that are not mentioned in the paper, so please choose your experimental settings in run.py carefully during the re-creation.
If you have questions or suggestions about this code, please ask in the issue and I will try to answer them.
If the paper and code have helped you, please give me a star!