Generative Adversarial Network for Abstractive Text Summarization

Linqing Liu, Yao Lu, Min Yang, Qiang Qu, Jia Zhu, Hongyan Li

AAAI 2018 pp. 8109-8110

doi:10.1609/AAAI.V32I1.12141 /aaai/2018/liu2018aaai-generative/

Abstract

In this paper, we propose an adversarial process for abstractive text summarization, in which we simultaneously train a generative model G and a discriminative model D. In particular, we build the generator G as an agent of reinforcement learning, which takes the raw text as input and predicts the abstractive summarization. We also build a discriminator which attempts to distinguish the generated summary from the ground truth summary. Extensive experiments demonstrate that our model achieves competitive ROUGE scores with the state-of-the-art methods on CNN/Daily Mail dataset. Qualitatively, we show that our model is able to generate more abstractive, readable and diverse summaries.

PDF AAAI Semantic Scholar

Cite

Text

Liu et al. "Generative Adversarial Network for Abstractive Text Summarization." AAAI Conference on Artificial Intelligence, 2018. doi:10.1609/AAAI.V32I1.12141

Markdown

[Liu et al. "Generative Adversarial Network for Abstractive Text Summarization." AAAI Conference on Artificial Intelligence, 2018.](https://mlanthology.org/aaai/2018/liu2018aaai-generative/) doi:10.1609/AAAI.V32I1.12141

BibTeX

@inproceedings{liu2018aaai-generative,
  title     = {{Generative Adversarial Network for Abstractive Text Summarization}},
  author    = {Liu, Linqing and Lu, Yao and Yang, Min and Qu, Qiang and Zhu, Jia and Li, Hongyan},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2018},
  pages     = {8109-8110},
  doi       = {10.1609/AAAI.V32I1.12141},
  url       = {https://mlanthology.org/aaai/2018/liu2018aaai-generative/}
}