News Articles Dataset with Summary

该数据集是从纽约时报、CNN、Business Insider 和 Breitbart 抓取的新闻文章。 Kaggle 发表的原始数据集没有提供任何人工摘要，它只提供了文章的标题，虽然这可以用作摘要，但标题标题太短并不理想。我们通过为可用文章添加人工摘要来手动生成标签。我们还在数据集中添加了另一个名为主题的列，该列将说明新闻文章的类型。

该数据集非常适合摘要，因为提供的新闻文章很长并且会花费大量时间来阅读它。因此，为数据集中的文章生成自动摘要是理想的选择。该数据集由 50,001 行数据组成。

This dataset consists of news articles scraped from The New York Times, CNN, Business Insider, and Breitbart. The original dataset published on Kaggle did not include any human-written summaries, only the article titles. While titles can serve as crude summaries, they are too short to be ideal. We manually generated labels by adding human-written summaries for the available articles. We also added another column named "topic" to the dataset, which indicates the category of each news article. This dataset is highly suitable for summarization tasks, as the provided news articles are lengthy and require significant time to read. Thus, generating automatic summaries for the articles in this dataset is an ideal application. This dataset contains a total of 50,001 rows of data.

搜集汇总

背景与挑战

背景概述

该数据集包含从纽约时报、CNN等媒体抓取的新闻文章，并添加了人工摘要和主题列，适用于自动摘要任务。数据集共有50,001行数据，旨在解决原始数据中摘要缺失的问题。