Uncovering Conspiratorial Narratives within Arabic Online Content
It addresses a gap in conspiracy theory research by focusing on Arabic content, which has traditionally been overlooked in favor of English-language or offline data, providing insights for understanding public discourse in the Arab world.
This study tackled the spread of conspiracy theories in Arabic digital spaces by analyzing online content using Named Entity Recognition and Topic Modeling, uncovering six distinct categories such as geopolitical and government cover-ups, and highlighting their deep embedding in social media discourse shaped by regional contexts.
This study investigates the spread of conspiracy theories in Arabic digital spaces through computational analysis of online content. By combining Named Entity Recognition and Topic Modeling techniques, specifically the Top2Vec algorithm, we analyze data from Arabic blogs and Facebook to identify and classify conspiratorial narratives. Our analysis uncovers six distinct categories: gender/feminist, geopolitical, government cover-ups, apocalyptic, Judeo-Masonic, and geoengineering. The research highlights how these narratives are deeply embedded in Arabic social media discourse, shaped by regional historical, cultural, and sociopolitical contexts. By applying advanced Natural Language Processing methods to Arabic content, this study addresses a gap in conspiracy theory research, which has traditionally focused on English-language content or offline data. The findings provide new insights into the manifestation and evolution of conspiracy theories in Arabic digital spaces, enhancing our understanding of their role in shaping public discourse in the Arab world.