We present deep communicating agents in an encoder-decoder architecture to address the challenges of representing a long document for abstractive summarization. With deep communicating agents, the task of encoding a long text is divided across multiple collaborating agents, each in charge of a subsection of the input text. These encoders are connected to a single decoder, trained end-to-end using reinforcement learning to generate a focused and coherent summary. Empirical results demonstrate that multiple communicating encoders lead to a higher quality summary compared to several strong baselines, including those based on a single encoder or multiple non-communicating encoders.
@article{arxiv.1803.10357,
title = {Deep Communicating Agents for Abstractive Summarization},
author = {Asli Celikyilmaz and Antoine Bosselut and Xiaodong He and Yejin Choi},
journal= {arXiv preprint arXiv:1803.10357},
year = {2018}
}