Cross-Platform and Cross-Lingual Dynamics of Wikipedia Sharing and Contribution

Abstract

Wikipedia is a key knowledge source, but the literature suggests that attention from social media platforms rarely leads to edits. This research examines this relationship using a 15-year dataset of links from three major social media platforms (Twitter, Reddit, and Discord) that contains the link-sharing history of multiple languages and edits of Wikipedia articles. We first characterize the shared articles in terms of their language co-occurrence and editing history. Then, we apply a staggered difference-in-differences design with unshared language editions as controls. The present work reveals several distinctive platform-specific patterns. Reddit favors developed articles in a Eurocentric sphere, while Twitter is more dynamic and engages earlier. Our analysis also confirms the findings of prior work, finding that the association between sharing and editing activity is not strong. However, we also observed that high-intensity sharing correlates with measurable article growth. This study thus contributes to the literature by refining the one-way street view with comprehensive characteristics using multi-platform and multi-lingual analysis and showing that attention and contribution are conditionally connected.