Controllable Abstractive Summarization

TL;DR

This paper introduces a neural abstractive summarization model that allows user preferences to control summary attributes like length and style, improving quality and customization over existing models.

Contribution

It presents a novel controllable summarization framework enabling user-specified attributes, outperforming state-of-the-art models on the CNN-Dailymail dataset.

Findings

Outperforms state-of-the-art in ROUGE-1 score (40.38 vs. 39.53)

Produces higher quality summaries according to human evaluation

Effectively incorporates user preferences into summarization process

Abstract

Current models for document summarization disregard user preferences such as the desired length, style, the entities that the user might be interested in, or how much of the document the user has already read. We present a neural summarization model with a simple but effective mechanism to enable users to specify these high level attributes in order to control the shape of the final summaries to better suit their needs. With user input, our system can produce high quality summaries that follow user preferences. Without user input, we set the control variables automatically. On the full text CNN-Dailymail dataset, we outperform state of the art abstractive systems (both in terms of F1-ROUGE1 40.38 vs. 39.53 and human evaluation).