Bot Impacts on Public Sentiment and Community Structures: Comparative Analysis of Three Elections in the Asia-Pacific

Abstract

Online disinformation has become a ubiquitous concern in elections worldwide. However, while extensively studied in Western contexts, scant work examines its prevalence and impacts in other geopolitical settings like the Asia-Pacific. This paper probes the influence of online bots on Twitter conversations surrounding recent elections in the Philippines, Indonesia, and Taiwan. Using a combination of machine learning, network analysis, and causal inference tools, we determine that the impacts of bots are mixed across contexts. More specifically, we quantify variations in the extent to which bot activities account for shifts in public sentiment and online community structure over time. We contribute to the extensive literature on online disinformation by providing a general and systematic framework for assessing and comparing the impacts of bot operations across unique geopolitical contexts.