Get our free extension to see links to code for papers anywhere online!

 Add to Chrome

 Add to Firefox

CatalyzeX Code Finder - Browser extension linking code for ML papers across the web! | Product Hunt Embed

Deep Communicating Agents for Abstractive Summarization

Aug 15, 2018
Asli Celikyilmaz, Antoine Bosselut, Xiaodong He, Yejin Choi



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.

* Accepted for publication at NAACL 2018 


Share this with someone who'll enjoy it:

   Access Paper Source



Share this with someone who'll enjoy it: