This page displays the main data used in the paper A Thousand Words Tell More Than Just Numbers: Financial Crises and Historical Headlines (2021) by Ristolainen, K., T. Roukka, H. Nyberg.
The topic frequencies are derived from the output of a Latend Dirichlei Allocation topic model that is estimated with news article titles covering economy related issues for 17 developed countries from 1870 to 2016.
The news article titles are from six global newspapers.
The news article data is from ProQuest Historical newspapers.
More detailed information about the data collection and model estimation process can be found from the paper.
