Enriching and Controlling Global Semantics for Text Summarization
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2021
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914555890761728 |
|---|---|
| author | Nguyen, Thong Luu, Anh Tuan Lu, Truc Quan, Tho |
| author_facet | Nguyen, Thong Luu, Anh Tuan Lu, Truc Quan, Tho |
| contents | Recently, Transformer-based models have been proven effective in the abstractive summarization task by creating fluent and informative summaries. Nevertheless, these models still suffer from the short-range dependency problem, causing them to produce summaries that miss the key points of document. In this paper, we attempt to address this issue by introducing a neural topic model empowered with normalizing flow to capture the global semantics of the document, which are then integrated into the summarization model. In addition, to avoid the overwhelming effect of global semantics on contextualized representation, we introduce a mechanism to control the amount of global semantics supplied to the text generation module. Our method outperforms state-of-the-art summarization models on five common text summarization datasets, namely CNN/DailyMail, XSum, Reddit TIFU, arXiv, and PubMed. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2109_10616 |
| institution | arXiv |
| publishDate | 2021 |
| record_format | arxiv |
| spellingShingle | Enriching and Controlling Global Semantics for Text Summarization Nguyen, Thong Luu, Anh Tuan Lu, Truc Quan, Tho Computation and Language Recently, Transformer-based models have been proven effective in the abstractive summarization task by creating fluent and informative summaries. Nevertheless, these models still suffer from the short-range dependency problem, causing them to produce summaries that miss the key points of document. In this paper, we attempt to address this issue by introducing a neural topic model empowered with normalizing flow to capture the global semantics of the document, which are then integrated into the summarization model. In addition, to avoid the overwhelming effect of global semantics on contextualized representation, we introduce a mechanism to control the amount of global semantics supplied to the text generation module. Our method outperforms state-of-the-art summarization models on five common text summarization datasets, namely CNN/DailyMail, XSum, Reddit TIFU, arXiv, and PubMed. |
| title | Enriching and Controlling Global Semantics for Text Summarization |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2109.10616 |