Dark Mode
Politifact Fact-Checked News
Entertainment & Media Consumption
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This dataset aims to address the critical issue of misinformation, which significantly impacts public perception and decision-making. It contains approximately 10,000 news articles and associated metadata, primarily scraped from the Politifact website. The dataset is designed to help data scientists analyse the spread of fake news and develop models to classify news articles as either false or true, contributing to efforts to combat the propagation of misleading information. It provides a valuable resource for understanding the characteristics of fact-checked news content.
Columns
- News_Headline: This column contains the textual content of the news information that requires analysis.
- Link_Of_News: Provides the URL linking to the original news article.
- Source: Identifies the authors or entities who posted the news information on various social media platforms, such as Facebook, Instagram, or Twitter.
- Stated_On: Indicates the date when the news information was initially posted by the source on social media.
- Date: Specifies the date when the Politifact fact-checking team verified and categorised the news information.
- Label: Contains the classification assigned to each news item. This column includes five distinct labels: True, Mostly-True, Half-True, Barely-True, False, and Pants on Fire. Users can choose to perform multi-class classification or convert these labels for binary classification (e.g., True or False).
Distribution
This dataset comprises approximately 10,000 individual news articles and their associated metadata, structured with six primary attributes. The data is typically provided in a CSV file format. The unique values for key attributes such as News_Headline, Link_Of_News, Stated_On, Date, and Label are consistently around 9,947 to 9,960 records, while the 'Source' column has 1,028 unique values.
Usage
This dataset is ideal for a range of analytical and machine learning applications. It can be used to gain insights into how to halt the spread of misinformation and to determine which approaches offer superior accuracy in combating fake news. Specific use cases include developing and training machine learning models for news classification, performing multi-class classification to distinguish between different degrees of truthfulness, or converting labels for binary classification (Fake vs. Real). It is particularly well-suited for projects involving Natural Language Processing (NLP) and Data-Mining concepts.
Coverage
The dataset's content is global in its relevance, as fake news is a worldwide concern. The information covers a time range from 20 June 2013 to 19 June 2020. The collection of data on different dates includes:
- 20 June 2013 - 02 March 2014: 839 records
- 02 March 2014 - 13 November 2014: 975 records
- 13 November 2014 - 26 July 2015: 857 records
- 26 July 2015 - 07 April 2016: 981 records
- 07 April 2016 - 19 December 2016: 1,286 records
- 19 December 2016 - 31 August 2017: 881 records
- 31 August 2017 - 14 May 2018: 873 records
- 14 May 2018 - 24 January 2019: 982 records
- 24 January 2019 - 07 October 2019: 956 records
- 07 October 2019 - 19 June 2020: 1,330 records The dataset includes news information posted by various sources on social media platforms and fact-checked by the Politifact.com team.
License
CC BY-SA
Who Can Use It
This dataset is primarily intended for data scientists who are interested in tackling the problem of misinformation. Users can leverage this data to train their machine learning models to identify and classify fake news, contributing to the broader effort to improve information accuracy. It supports research and development in areas such as natural language processing, data mining, and automated fact-checking.
Dataset Name Suggestions
- Fake-Real News Dataset
- Politifact Fact-Checked News
- Misinformation Detection Corpus
- Social Media News Verification Dataset
- News Authenticity Classifier Data
Attributes
Original Data Source: Fake-Real News
QUALITY