Deciphering Viral Trends in WhatsApp: A Case Study From a Village in Rural India (2407.08172v1)

Abstract: This research studies the nature and spread of WhatsApp content among everyday users in a rural Indian village. Leveraging a dataset of hundreds of private WhatsApp groups collected with consent from participants, our study uncovers the kinds of WhatsApp groups users are part of, marking the first such categorization. The dataset comprises tens of thousands of messages, out of which we manually classified 604 pieces of content designated as 'forwarded many times'-indicating their viral status. Our key findings indicate a high prevalence of content focused on national politics, with the viral messages overwhelmingly supporting a specific political party and disparaging the opposition. Significantly, these messages were fraught with misinformation, engendering hate against Muslims and promoting a narrative of Hindus being under threat. This trend was particularly noticeable within caste-based groups, which were dominated by misinformation, pro-BJP rhetoric, anti-Congress content, and Hindutva propaganda. Remarkably, much of the misinformation circulating had previously been discredited by established fact-checking organizations. This suggests not only a recurring cycle of debunked information reappearing but also that fact-checks are failing to penetrate these specific groups. As the first quantitative analysis of everyday WhatsApp use in a rural context, this research has far-reaching implications for understanding the unique challenges posed by end-to-end encrypted platforms. It serves as a crucial baseline for designing more effective moderation policies aimed at combating misinformation and fostering a more responsible use of encrypted communication channels.