Abstract

This study examines whether elections exacerbate intergroup conflicts by using text mining techniques. To this end, we collected social media data before and after the 2024 South Korean parliamentary election and analyzed changes in hate speech targeting key electoral conflict factors: age, gender, and region. Beta regression analysis revealed that hate speech against these groups significantly increased during the election year (2024) and on election day (April 10). Notably, on election day within the election year, hate speech targeting age, region, and gender increased, whereas no such change was observed in the non-election year. These findings suggest that elections, often referred to as the “flower of democracy,” may be associated with a rise in online hate speech. This study provides a foundational basis for developing intervention strategies to mitigate the spread of hate speech during election periods.