Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2305.06647 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914693258412032 |
|---|---|
| author | Ma, Xinbei Gong, Yeyun He, Pengcheng Zhao, Hai Duan, Nan |
| author_facet | Ma, Xinbei Gong, Yeyun He, Pengcheng Zhao, Hai Duan, Nan |
| contents | Based on the remarkable achievements of pre-trained language models in abstractive summarization, the copying mechanism has proved helpful by improving the factuality, stability, and overall performance. This work proposes PROM, a new PhRase-level cOpying Mechanism that enhances attention on n-grams, which can be applied to zero-shot summarization with pre-training. PROM adds an indicator layer to explicitly pick up tokens in n-gram that can be copied from the source, and calculates an auxiliary loss for the copying prediction. Empirical studies show that PROM makes significant improvements in fine-tuning on benchmarks. In zero-shot setting, PROM is utilized in the self-supervised pre-training on raw corpora and provides new general baselines on a wide range of summarization datasets. Further analysis shows that PROM performs more reasonable copying and contributes to faithfulness. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2305_06647 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | PROM: A Phrase-level Copying Mechanism with Pre-training for Abstractive Summarization Ma, Xinbei Gong, Yeyun He, Pengcheng Zhao, Hai Duan, Nan Computation and Language Based on the remarkable achievements of pre-trained language models in abstractive summarization, the copying mechanism has proved helpful by improving the factuality, stability, and overall performance. This work proposes PROM, a new PhRase-level cOpying Mechanism that enhances attention on n-grams, which can be applied to zero-shot summarization with pre-training. PROM adds an indicator layer to explicitly pick up tokens in n-gram that can be copied from the source, and calculates an auxiliary loss for the copying prediction. Empirical studies show that PROM makes significant improvements in fine-tuning on benchmarks. In zero-shot setting, PROM is utilized in the self-supervised pre-training on raw corpora and provides new general baselines on a wide range of summarization datasets. Further analysis shows that PROM performs more reasonable copying and contributes to faithfulness. |
| title | PROM: A Phrase-level Copying Mechanism with Pre-training for Abstractive Summarization |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2305.06647 |