Highlighting Entanglement of Cultures via Ranking of Multilingual Wikipedia Articles
This work addresses the problem of quantifying cultural diversity and shared knowledge for researchers in computational social science, though it is incremental as it applies existing ranking methods to new multilingual data.
The study tackled the problem of understanding cultural differences in evaluating important persons by ranking multilingual Wikipedia articles using three network-based algorithms, finding that while local heroes dominate, global heroes also exist, forming an effective network that reveals cultural entanglements and follows a Zipf law distribution.
How different cultures evaluate a person? Is an important person in one culture is also important in the other culture? We address these questions via ranking of multilingual Wikipedia articles. With three ranking algorithms based on network structure of Wikipedia, we assign ranking to all articles in 9 multilingual editions of Wikipedia and investigate general ranking structure of PageRank, CheiRank and 2DRank. In particular, we focus on articles related to persons, identify top 30 persons for each rank among different editions and analyze distinctions of their distributions over activity fields such as politics, art, science, religion, sport for each edition. We find that local heroes are dominant but also global heroes exist and create an effective network representing entanglement of cultures. The Google matrix analysis of network of cultures shows signs of the Zipf law distribution. This approach allows to examine diversity and shared characteristics of knowledge organization between cultures. The developed computational, data driven approach highlights cultural interconnections in a new perspective.