MAPPING RESEARCH TRENDS ON POLITICAL POLARIZATION IN ONLINE MEDIA: A BIBLIOMETRIC ANALYSIS USING VOSVIEWER

Abstract

Political polarization in online media has become a prominent research theme, especially around elections, yet its intellectual structure remains fragmented across disciplines. This study maps the scholarly landscape of research linking online media, political polarization, and election through a bibliometric analysis of Scopus-indexed publications, visualized with VOSviewer, and complemented by a narrative synthesis of Discourse Network Analysis (DNA) studies. A Scopus search using the combined keywords “online media,” “political polarization,” and “election” retrieved 198 documents, processed into a keyword co-occurrence network of 59 items grouped into four clusters. The results show that “social media polarization,” “political polarization,” and “sentiment analysis” are the most dominant and interconnected terms, indicating that existing research concentrates on social media platforms and computational methods such as sentiment analysis and machine learning, while studies of polarization in verified online news media at the local-election level remain scarce. A narrative synthesis of sixteen studies (2020–2026) further shows that DNA is increasingly applied to Indonesian public debates but has rarely been extended to local electoral contests such as regional head elections (Pilkada). The novelty of this study lies in combining bibliometric mapping with narrative synthesis to demonstrate that discourse-network research on polarization in verified online news coverage of local elections constitutes an underexplored, and within the databases searched apparently unoccupied, gap in the literature. These findings offer a methodological and thematic roadmap for future communication research on media-driven political polarization at the sub-national level.