Topic detection based on weak tie analysis: A case study of LIS research

Abstract

Purpose: Based on the weak tie theory, this paper proposes a series of connection indicators of weak tie subnets and weak tie nodes to detect research topics, recognize their connections, and understand their evolution. Design/methodology/approach: First, keywords are extracted from article titles and preprocessed. Second, high-frequency keywords are selected to generate weak tie co-occurrence networks. By removing the internal lines of clustered sub-topic networks, we focus on the analysis of weak tie subnets' composition and functions and the weak tie nodes' roles. Findings: The research topics' clusters and themes changed yearly; the subnets clustered with technique-related and methodology-related topics have been the core, important subnets for years; while close subnets are highly independent, research topics are generally concentrated and most topics are application-related; the roles and functions of nodes and weak ties are diversified. Research limitations: The parameter values are somewhat inconsistent; the weak tie subnets and nodes are classified based on empirical observations, and the conclusions are not verified or compared to other methods. Practical implications: The research is valuable for detecting important research topics as well as their roles, interrelations, and evolution trends. Originality/value: To contribute to the strength of weak tie theory, the research translates weak and strong ties concepts to co-occurrence strength, and analyzes weak ties' functions. Also, the research proposes a quantitative method to classify and measure the topics' clusters and nodes.