Reinforced Extractive Summarization with Question-Focused Rewards

By Kristjan Arumae, Fei Liu. 2018

33 citations recorded by the source corpus.

We investigate a new training paradigm for extractive summarization. Traditionally, human abstracts are used to derive goldstandard labels for extraction units. However, the labels are often inaccurate, because human abstracts and source documents cannot be easily aligned at the word level. In this paper we convert human abstracts to a set of Cloze-style comprehension questions. System summaries are encouraged to preserve salient source content useful for answering questions and share common words with the abstracts. We use reinforcement learning to explore the space of possible extractive summaries and introduce a question-focused reward function to promote concise, fluent, and informative summaries. Our experiments show that the proposed method is effective. It surpasses state-of-the-art systems on the standard summarization dataset.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

Finding Generalizable Evidence by Learning to Convince Q&A Models (2019, background): “…artificial-intelligence/project-debater Generic Summarization In contrast, various papers focus primarily on summarization rather than QA, using downstream QA accuracy only as a reward to optimize generic (question-agnostic) summarization models (Arumae and Liu, 2018, 2019; Eyal et al., 2019).”
Guiding Extractive Summarization with Question-Answering Rewards (2019, methodology): “Following Arumae and Liu (2018), we set hyperparameters β = 2 α ; α and γ are tuned on the dev set using grid search.”
Searching for Effective Neural Extractive Summarization: What Works and What’s Next (2019, methodology): “, 2018) and learning schemas to introduce extra instructive constraints (Paulus et al., 2017; Arumae and Liu, 2018).”
Extractive Summarization as Text Matching (2020, methodology): “…encoder-decoder framework by choosing RNN (Zhou et al., 2018), Transformer (Zhong et al., 2019b; Wang et al., 2019) or GNN (Wang et al., 2020) as encoder, non-auto-regressive (Narayan et al., 2018b; Arumae and Liu, 2018) or auto-regressive decoders (Jadhav and Rajan, 2018; Liu and Lapata, 2019).”
Heterogeneous Graph Neural Networks for Extractive Document Summarization (2020, background): “Recent years have seen a resounding success in the use of deep neural networks on this task (Cheng and Lapata, 2016; Narayan et al., 2018; Arumae and Liu, 2018; Zhong et al., 2019a; Liu and Lapata, 2019b).”
The Summary Loop: Learning to Write Abstractive Summaries Without Examples (2020, background): “Summarization and Q&A. Eyal et al. (2019) and Arumae and Liu (2018) turn reference summaries into ﬁll-in-the-blank (FIB) questions, either as an evaluation metric or to train an extractive summarization model.”