Deep reinforcement learning (RL) has been a commonly-used strategy for the abstractive summarization task to address both the exposure bias and non-differentiable task issues. However, the conventional reward Rouge-L simply looks for exact n-grams matches between candidates and annotated references, which inevitably makes the generated sentences repetitive and incoherent. In this paper, instead of Rouge-L, we explore the practicability of utilizing the distributional semantics to measure the matching degrees. With distributional semantics, sentence-level evaluation can be obtained, and semantically-correct phrases can also be generated without being limited to the surface form of the reference sentences. Human judgments on Gigaword and CNN/Daily Mail datasets show that our proposed distributional semantics reward (DSR) has distinct superiority in capturing the lexical and compositional diversity of natural language.
@article{arxiv.1909.00141,
title = {Deep Reinforcement Learning with Distributional Semantic Rewards for Abstractive Summarization},
author = {Siyao Li and Deren Lei and Pengda Qin and William Yang Wang},
journal= {arXiv preprint arXiv:1909.00141},
year = {2019}
}