Informing policy with text mining: technological change and social challenges

Abstract This study presents an innovative text mining methodology that supports policy analysts with problem recognition, definition and selection. The empirical analysis is based on four years of online news articles published in the period 2016–2019. Using a combination methods (analysis term-frequencies, co-occurrence sentiment analysis), we identify trending terms explore selected regulatory issues. demonstrates while each algorithm provides insightful results, their yields more detailed robust overview problems. results present early signals trends, connections between topics, changing public attitudes towards them.