Identifying influential nodes of global terrorism network: A comparison for skeleton network extraction

• Main Authors: Kanokwan Malang, Shuliang Wang, Aniwat Phaphuangwittayakul, Yuanyuan Lv, Hanning Yuan, Xiuzhen Zhang
• Format: Journal
• Published: 2020
• Subjects: Mathematics; Physics and Astronomy
• Online Access: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85077700485&origin=inward; http://cmuir.cmu.ac.th/jspui/handle/6653943832/68450
• Institution: Chiang Mai University

© 2019 Elsevier B.V. The inherent structure and substantial information on global terrorism network are often understood by identifying influential nodes. Recently, novel node identification methods are developed from different perspectives. Each of them has trade-offs and strengths. However, the algorithms for exploring the key influential nodes have been adopted unevenly in light of network extraction research. A set of nodes that is more favorable to define the core network structure is unclear. In this paper, we, therefore, present a comparative study of node identification methods over the global terrorism network. The new insight each method contributes to identifying key influential nodes and core network structure is investigated. Six comparative methods are verified by the SIR model and monotonicity index. We further elaborate on experimental analysis by applying the critical nodes from each method to extract the skeleton network. All extracted skeletons are eventually compared with the original network in terms of node correlation and network structural-equivalence. Thus, the comparison and results not only used to reflect the potential of different methods to a particular network structure but also guide us to select a method that works best for extracting the skeleton network of real-world global terrorism.