Domain-relevance of influence: characterizing variations in online influence across multiple domains on social media

Abstract

AbstractInfluentials play a key role in enhancing information diffusion on social media. However, how personal influence varies across multiple domains is rarely addressed. This study introduces a concept called Domain-Relevance of Influence to describe the relation between influence and domains, and establishes a methodological framework with a sample of 8,520,933 Weibo users to explore the cross-domain characteristics of influence. The results show that generalists with cross-domain attributes possess significantly higher influence than specialists in most domains, whereas in a single domain such as sports or technology, specialists and generalists can possess comparable influence. We further show that influence is positively associated with cross-domain capability in overall domains, but not necessarily in each single domain. This study contributes to better understanding of the influence variation across domains for influence enhancement, and provides a big data-based methodological basis for cross-domain communication research.