[go: up one dir, main page]

Published October 1, 2020 | Version v1
Dataset Open

Webis-News-Bias-20

  • 1. Paderborn University
  • 2. Bauhaus-Universität Weimar

Description

This is the dataset used in the paper "Analyzing Political Bias and Unfairness in News Articles at Different Levels of Granularity."

The news articles are stored in json format, one news each line. For each news, there are following fields:

 - title: the title of the news
 - content: the content of the news
 - source: the news portal
 - allsides_bias: the bias indicated in the allsides.com (left, center, or right)
 - misc: other information, such as author, date, and topics
 - adfontes_fair: the labels from adfontesmedia.com, whehter the article is fair or not (bias, neutral, or unknown)
 - adfontes_political: the labels from adfontesmedia.com, whether the article is political bias or not (bias, neutral, or unknown)
 - event_id: the event id. Articles with the same event have the same id

 

Files

README.md

Files (13.8 MB)

Name Size Download all
md5:5db79856cc90bc2a823684a0c4e17ed8
874 Bytes Preview Download
md5:5fa39c7088cd7f239414f6c90ea0bbc2
13.8 MB Preview Download

Additional details

Related works