Summary Level Training of Sentence Rewriting for Abstractive Summarization

Abstract

As an attempt to combine extractive and abstractive summarization, Sentence Rewriting models adopt the strategy of extracting salient sentences from a document first and then paraphrasing the selected ones to generate a summary. However, the existing models in this framework mostly rely on sentence-level rewards or suboptimal labels, causing a mismatch between a training objective and evaluation metric. In this paper, we present a novel training signal that directly maximizes summary-level ROUGE scores through reinforcement learning. In addition, we incorporate BERT into our model, making good use of its ability on natural language understanding. In extensive experiments, we show that a combination of our proposed model and training procedure obtains new state-of-the-art performance on both CNN/Daily Mail and New York Times datasets. We also demonstrate that it generalizes better on DUC-2002 test set.

Neural Extractive Summarization with Hierarchical Attentive Heterogeneous Graph Network

“…These work usually instantiate their encoder-decoder architecture by choosing RNN (Nallapati et al, 2017;Zhou et al, 2018), Transformer Zhong et al, 2019b;Liu and Lapata, 2019;Zhang et al, 2019b) or Hierarchical GNN as encoder, autoregressive (Jadhav and Rajan, 2018;Liu and Lapata, 2019) or nonautoregressive (Narayan et al, 2018;Arumae and Liu, 2018) decoders. The application of RL provides a means of summary-level scoring and brings improvement (Narayan et al, 2018;Bae et al, 2019).…”

“…As we all know, the autoregressive paradigm faces error propagation and exposure bias problems (Ranzato et al, 2015). Besides, reinforcement learning is also introduced to consider the semantics of extracted summary (Narayan et al, 2018;Bae et al, 2019), which combines the maximum-likelihood cross-entropy loss with the rewards from policy gradient to directly optimize the evaluation metric for the summarization task. Recently, the popular solution is to build a summarization system with two-stage decoder.…”

“…Recently, the popular solution is to build a summarization system with two-stage decoder. These models extract salient sentences and then rewrite (Chen and Bansal, 2018;Bae et al, 2019), compress (Lebanoff et al, 2019;Xu and Durrett, 2019;Mendes et al, 2019), or match (Zhong et al, 2020) these sentences.…”

Fact-level Extractive Summarization with Hierarchical Graph Mask on BERT

“…proposes a decoder that jointly learns to score and select sentences, while (Zhang et al, 2018) presents a latent variable extractive summarization model, which directly maximizes the likelihood of human summaries. In recent works, models based on pretrained models Bae et al, 2019;Zhang et al, 2019), especially BERT, have made a step forward.…”

“…Recent works Bae et al, 2019;Zhang et al, 2019) have demonstrated that it is highly beneficial to apply pretrained language models such as BERT to extractive summarization models. Following this trend, we adopt a BERT-based encoder to generate contextualized representations for further extraction.…”