Multi-period configurational patterns for emotion polarization regulation during major catastrophic emergencies

Abstract

Major catastrophic emergencies often give rise to a polarization of negative public sentiment within public opinion. Without proper regulation, such polarization can instigate public fear, anger, or even erode government credibility. This paper constructs configurational patterns for official media’s emotion polarization regulation during major emergencies from a causal configuration perspective. Multi-stage nature of major emergencies and the multiple types of public opinion information are considered. Using the China Eastern Airlines MU5735 crash as a study case, trending data throughout the entire cycle of this emergency from Sina Weibo was collected via web crawlers and categorized through topic clustering. This paper defined positive sentiment entropy and constructed sentiment dictionary based on this case. Based on the Yale attitude change model, fifteen potential conditions were specified. To mitigate data redundancy, ensemble models were used to identify key conditions. Multi-period fuzzy-set qualitative comparative analysis (mfsQCA) was then used to identify configurational patterns for moderating collective emotion polarization and deriving decision-oriented implications. This paper offers a configurational perspective on regulating emotion polarization in online public opinion. It also contributes to crisis communication and disaster management by offering methodological insights for moderating public sentiment during emergencies.