ACL Findings2024
GENDEX: Generative Data Augmentation Strategy Leveraging External Data for Abstractive Dialogue Summarization
Sangwon Park, Hongseok Choi, Dongha Choi, Hyunju Lee
摘要
With the proliferation of digital communication, dialogue summarization has become increasingly important. However, it still faces a shortage of data. To address this issue, we developed Generative Data Augmentation Strategy Leveraging External Data for Abstractive Dialogue Summarization (GENDEX), which is based on the hypothetical foundation that texts containing people and their interpersonal interactions can potentially serve as summaries of corresponding dialogues. We filter short texts containing people and resolve coreferences for better contextual analysis. We then identify the semantic roles of words within the texts and filter them based on the patterns observed in the dialogue summarization datasets. Using these texts, we generate synthetic dialogues through a controlled generation method. To better leverage the augmented data, we utilize noisetolerant training to fine-tune the summarization model. The experimental results demonstrate the effectiveness of our proposed method, showing its robust performance, generalizability, and scalability. Moreover, performance improvements by GENDEX were observed regardless of complexity of dialogues. The code is available at https://github.com/DMCB- GIST/GENDEX.
问 Lune
几分钟读透这篇 Paper,不用再花一下午。
Lune 能读完全文,也能读引用它的顶会 Paper,讲清它的结论、证据和局限,每一点都注明出自哪篇 Paper。不用硬啃,也不靠猜。
脉络
- An Exploratory Study on Long Dialogue Summarization: What Works and What's Next 引用了 Multi-View Sequence-to-Sequence Models with Conversational Structure for Abstractive Dialogue Summarization
- DialogSum: A Real-Life Scenario Dialogue Summarization Dataset 引用了 Multi-View Sequence-to-Sequence Models with Conversational Structure for Abstractive Dialogue Summarization
它引用的 Paper8
- PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive SummarizationJingqing Zhang, Yao Zhao, Mohammad Saleh, Peter J. LiuICML 2020 · 2,453 次引用
- DialogSum: A Real-Life Scenario Dialogue Summarization DatasetYulong Chen, Yang Liu, Liang Chen, Yue ZhangACL Findings 2021 · 157 次引用
- Multi-View Sequence-to-Sequence Models with Conversational Structure for Abstractive Dialogue SummarizationJiaao Chen, Diyi YangEMNLP 2020 · 121 次引用
- Towards a Unified Multi-Dimensional Evaluator for Text GenerationMing Zhong, Yang Liu, Da Yin, Yuning Mao 等EMNLP 2022 · 103 次引用
- An Exploratory Study on Long Dialogue Summarization: What Works and What's NextYusen Zhang, Ansong Ni, Tao Yu, Rui Zhang 等EMNLP Findings 2021 · 49 次引用
相关 Paper
- Simple Conversational Data Augmentation for Semi-supervised Abstractive Dialogue SummarizationJiaao Chen, Diyi YangEMNLP 2021 · 31 次引用
- Compositional Data Augmentation for Abstractive Conversation SummarizationSiru Ouyang, Jiaao Chen, Jiawei Han, Diyi YangACL 2023 · 4 次引用
- Controllable Abstractive Dialogue Summarization with Sketch SupervisionChien-Sheng Wu, Linqing Liu, Wenhao Liu, Pontus Stenetorp 等ACL Findings 2021 · 47 次引用
- Low-Resource Dialogue Summarization with Domain-Agnostic Multi-Source PretrainingYicheng Zou, Bolin Zhu, Xingwu Hu, Tao Gui 等EMNLP 2021 · 18 次引用
- Data Augmentation for Low-Resource Keyphrase GenerationKrishna Garg, Jishnu Ray Chowdhury, Cornelia CarageaACL Findings 2023 · 8 次引用
