End-to-End Segmentation-based News Summarization

TL;DR

This paper introduces a new task of segmenting news articles into sections and generating summaries for each, supported by a new dataset and a novel segmentation-based language model that outperforms existing models.

Contribution

The paper presents the SegNews dataset and a novel segmentation-based language model for section-wise news summarization, advancing the capabilities of news content digestion.

Findings

Model outperforms state-of-the-art sequence-to-sequence models

SegNews dataset contains 27k articles with section summaries

Joint segmentation and summarization improves content understanding

Abstract

In this paper, we bring a new way of digesting news content by introducing the task of segmenting a news article into multiple sections and generating the corresponding summary to each section. We make two contributions towards this new task. First, we create and make available a dataset, SegNews, consisting of 27k news articles with sections and aligned heading-style section summaries. Second, we propose a novel segmentation-based language generation model adapted from pre-trained language models that can jointly segment a document and produce the summary for each section. Experimental results on SegNews demonstrate that our model can outperform several state-of-the-art sequence-to-sequence generation models for this new task.