Reinforcing Semantic-Symmetry for Document Summarization

TL;DR

This paper introduces ReSyM, a reinforcement learning-based model that explicitly addresses semantic gaps in document summarization, leading to more semantically consistent summaries and outperforming state-of-the-art methods on benchmark datasets.

Contribution

ReSyM is the first to explicitly incorporate semantic consistency rewards for both extractor and abstractor modules in document summarization.

Findings

ReSyM outperforms baselines on CNN/Daily Mail and BigPatent datasets.

Semantic rewards improve summary semantic fidelity.

Enhanced sentence representation learning boosts performance.

Abstract

Document summarization condenses a long document into a short version with salient information and accurate semantic descriptions. The main issue is how to make the output summary semantically consistent with the input document. To reach this goal, recently, researchers have focused on supervised end-to-end hybrid approaches, which contain an extractor module and abstractor module. Among them, the extractor identifies the salient sentences from the input document, and the abstractor generates a summary from the salient sentences. This model successfully keeps the consistency between the generated summary and the reference summary via various strategies (e.g., reinforcement learning). There are two semantic gaps when training the hybrid model (one is between document and extracted sentences, and the other is between extracted sentences and summary). However, they are not explicitly considered in the existing methods, which usually results in a semantic bias of summary. To mitigate the above issue, in this paper, a new \textbf{r}einforcing s\textbf{e}mantic-\textbf{sy}mmetry learning \textbf{m}odel is proposed for document summarization (\textbf{ReSyM}). ReSyM introduces a semantic-consistency reward in the extractor to bridge the first gap. A semantic dual-reward is designed to bridge the second gap in the abstractor. The whole document summarization process is implemented via reinforcement learning with a hybrid reward mechanism (combining the above two rewards). Moreover, a comprehensive sentence representation learning method is presented to sufficiently capture the information from the original document. A series of experiments have been conducted on two wildly used benchmark datasets CNN/Daily Mail and BigPatent. The results have shown the superiority of ReSyM by comparing it with the state-of-the-art baselines in terms of various evaluation metrics.