The ASSENZA-HUBER FAKE NEWS ATLAS Database is the result of an extensive data collection effort. This website provides freely available (open access) under clearly defined Terms of Use and License Terms, see below. We explicitly forbid commercial data providers from integrating (part of) our data into their services to generate profits. The aim of the Fake News Atlas is to provide a unified, open-access, and user-friendly database of fact-checked news to promote research across disciplines for whom fake news is relevant. To this end, we will provide the following information for each fact-checked news item. 

• First, the basics: the news, the publication date, and the news source where it was first published. 

• Second, we will provide keywords that help classify the news. 

• Third, we will provide a clear overall classification of the topic (e.g., public health, politics, economy, finance, etc.). 

• Fourth, for news items that fall under the category “economy”, we will classify each news item by the various subtopics (e.g., technology, taxes, government finance – budget, deficit, debt – labor markets, financial markets, inflation and prices). 

• Lastly, we will classify each news item by “emotion and sentiment”. We rely on sentiment analysis, which is a branch of natural language processing that focuses on detecting and understanding the emotional tone of a piece of text. Sentiment analysis, therefore, aims to identify whether a statement is expressing positive, negative, or neutral sentiment towards a particular topic. We rely on a specific tool, FinVADER, developed by Petr Korab, which is an updated version of the VADER (Valence Aware Dictionary and sEntiment Reasoner) classifier - a mainstream model for sentiment analysis using a general-language human-curated lexicon (see Hutto and Gilbert, 2014). 

The ultimate goal is to provide a unified, open-access, and user-friendly database of fact-checked news in English, French, German, and Italian to promote research across disciplines for whom fake news is relevant. We envision a data collection effort over several years. The database will bring together fact-checked news data currently dispersed across various sources. 

HOW TO CITE

Under the terms of use, any information taken directly or indirectly from the Assenza-Huber Fake News Atlas should be cited as:

• Assenza, T, F Collard, P Feve and S Huber (2024), "From Buzz to Bust: How Fake News Shapes the Business Cycle", CEPR Discussion Paper No. 18912. CEPR Press, Paris & London. https://cepr.org/publications/dp18912

DOWNLOAD AND DOCUMENTATION

Users may download the entire dataset in EXCEL or STATA format.

The first release (R.1) is coming soon. R.1 will cover the fact-checked news items used in our working paper "From Buzz to Bust: How Fake News Shapes the Business Cycle". The news items (published in English) have been fact-checked by a non-profit fact-checking organization adhering to the principles of the International Fact-Checking Network. 

More details about the data construction will be available soon here: pdf.

CONTACT

For questions, comments or suggestions about the data or webpage, please contact us at fakenewsatlas@gmail.com.

TERMS OF USE AND LICENSE TERMS

On this website, we provide open access under a license to an extensive fact-checked news dataset. Tiziana Assenza acknowledges funding from the French National Research Agency (ANR) under the “Investissements d’Avenir” program, grant ANR-17-EURE-0010. Stefanie Huber gratefully acknowledges support from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany ́s Excellence Strategy – EXC2126/1-390838866. We also thank our home institutions where we have conducted our research. The Assenza-Huber Fake News Atlas dataset is freely available in this non-commercial form.

Every user can use and/or share the licensed material, in whole or in part, provided that it is for non-commercial (e.g., academic) purposes, provided that our dataset is properly attributed and cited to credit the authors, and provided that it may only be shared under identical license terms. Commercial data providers are thus strictly forbidden to integrate all or parts of the dataset into their services and/or resell the data.

To comply with the attribution requirement in the license, whenever (part of) the dataset is used, it must be cited as follows: 

• Assenza, T, F Collard, P Feve and S Huber (2024), "From Buzz to Bust: How Fake News Shapes the Business Cycle", CEPR Discussion Paper No. 18912. CEPR Press, Paris & London. https://cepr.org/publications/dp18912

We advise making explicit reference to the date when the database was consulted, as statistics are subject to revisions. 

RESEARCH OUTPUT USING THIS DATABASE

NEWS 

1. New CEPR Discussion  Paper (March, 2024) "From Buzz to Bust: How Fake News Shapes the Business Cycle" (T. Assenza, F. Collard, P. Fève and S. J. Huber)

2. Non-technical Summary: VoxEU Column (April 10, 2024) "From Buzz to Bust: How Fake News Shapes the Business Cycle" (T. Assenza, F. Collard, P. Fève and S. J. Huber)