Newsroom: A Dataset of 1.3 Million Summaries with Diverse Extractive Strategies

Max Grusky, Mor Naaman, Yoav Artzi
4/30/2018
cs.CL

Official code: lil-lab/newsroom (151 stars, last commit 2023-08-08).

Abstract

We present NEWSROOM, a summarization dataset of 1.3 million articles and summaries written by authors and editors in newsrooms of 38 major news publications. Extracted from search and social media metadata between 1998 and 2017, these high-quality summaries demonstrate high diversity of summarization styles. In particular, the summaries combine abstractive and extractive strategies, borrowing words and phrases from articles at varying rates. We analyze the extraction strategies used in NEWSROOM summaries against other datasets to quantify the diversity and difficulty of our new data, and train existing methods on the data to evaluate its utility and challenges.

View on arXivView PDF

Code Implementations(1)

lil-lab/newsroomOfficial100%

Tools for downloading and analyzing summaries and evaluating summarization systems. https://summari.es/

15125HTML, PerlApr 27, 20183 years agoNOASSERTION

Similar Papers

Content Selection in Deep Learning Models of Summarization

83% match1 repo1062018

HunSum-1: an Abstractive Summarization Dataset for Hungarian

83% match1 repo112023

Neural Extractive Summarization with Side Information

83% match1 repo572017

Embrace Divergence for Richer Insights: A Multi-document Summarization Benchmark and a Case Study on Summarizing Diverse Information from News Articles

83% match1 repo122023

Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarization

81% match1 repo3782018

Analysis of Multidomain Abstractive Summarization Using Salience Allocation

81% match2024

Cite this paper

@article{grusky2018newsroom,
  title  = {Newsroom: A Dataset of 1.3 Million Summaries with Diverse Extractive Strategies},
  author = {Max Grusky and Mor Naaman and Yoav Artzi},
  year   = {2018},
  eprint = {1804.11283},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1804.11283}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1804.11283.svg)](https://codeofpaper.com/paper/1804.11283)

Discussion