SummaRuNNer: A Recurrent Neural Network based Sequence Model for Extractive Summarization of Documents

TL;DR

SummaRuNNer is an interpretable RNN-based model for extractive document summarization that outperforms or matches state-of-the-art methods and can be trained using only reference summaries.

Contribution

It introduces a novel RNN sequence model for extractive summarization with interpretability and a training approach that uses only human summaries, removing the need for sentence labels.

Findings

Achieves better or comparable performance to state-of-the-art methods.

Provides visualization of prediction features like salience and novelty.

Can be trained solely on human-generated summaries.

Abstract

We present SummaRuNNer, a Recurrent Neural Network (RNN) based sequence model for extractive summarization of documents and show that it achieves performance better than or comparable to state-of-the-art. Our model has the additional advantage of being very interpretable, since it allows visualization of its predictions broken up by abstract features such as information content, salience and novelty. Another novel contribution of our work is abstractive training of our extractive model that can train on human generated reference summaries alone, eliminating the need for sentence-level extractive labels.