The Interplay of Heuristic and Systematic Processing of Information in News Validation: A Natural Language Processing Approach

Abstract

With the growth of social media, user-generated content has become one of the key sources for news, but verifying its validity remains a challenge for both platforms and users. This study leverages the Heuristic-Systematic Model and in a novel approach employs state-of-the-art Natural Language Processing techniques to develop text features explaining the validity of online news from the user’s perspective. We used a panel dataset of over 130,000 labelled tweets to statistically validate our research model. Our results show that content similarity and loaded language (heuristic features), directly affect the perceived validity of online news. We also demonstrated the direct effect of language intensity and conflicting sentiment (systematic features) and their moderating effect of loaded language on the perceived validity of online news. Our proposed method sheds light on how platforms should analyze text features to evaluate online news credibility, which in turn provides a more reliable online environment.