Predicting Influential Cross-lingual Information Cascades on Twitter

Social network services (SNSs) have become new global and multilingual information platforms due to their popularity. In SNSs with content-sharing functionality, such as "retweets'' in Twitter and "share'' in Facebook, posts are easily and quickly shared among users, and some grow into large information cascades. Accompanied with such growth, cascades can spread over regions and languages. The "ALS Ice Bucket Challenge'' is a good example of internationally influential information that are widely spread and internationally reshared. Analyzing and predicting these influential cross-lingual information cascades will help promote culture exchange and detect international news and issues. This type of research can also provide insights into internal and external relationships among countries and languages. Though there has been a large amount of research on information cascades, much has been focused on predicting just their growth, and little has been on cross-region/lingual information cascades. In this work, we aim to build a robust model and detect influential cross-lingual information diffusion on social networks. To the first attempt, we analyze and predict influential cross-lingual information cascades on Twitter. With a large Twitter dataset, we conducted statistical analysis of growth and language distribution of information cascades. Based on the analysis, we propose a feature-based model to detect influential cross-lingual information cascades and successfully predict their size and language distribution. Keyword Information cascades; cross-lingual cascades; cascade growth; multilingualism