Automated Fact-Checking to Support Professional Practices: Systematic Literature Review and Meta-Analysis

Laurence Dierickx, Carl-Gustav Lindén, Andreas Lothe Opdahl

Abstract


Fact-checking is a time-consuming process that automation can potentially make more efficient. This study provides a comprehensive, multidisciplinary state of the art that considers a holistic and sociotechnical approach to studying automated fact-checking (AFC) from a journalistic perspective. It identifies how AFC tools, as boundary objects, connect with their end users. The findings highlight that most research in AFC focuses on providing technological solutions to solve the multidimensional social problem of information disorder, and they also indicate the promise of human-machine teaming. This suggests that the cursor can be moved from a technological point of view toward a social one, provided that a relationship of confidence is established between the communities—developers and fact-checkers/journalists—involved from either side of the tool. Although the conditions of use of a technological artifact are multifaceted, the reliability of the results that it provides participates in building such a trust-based relationship.


Keywords


fact-checking, journalism, artificial intelligence, systematic literature review, meta-analysis

Full Text:

PDF