Abstract
The identification of influential nodes in complex networks has been widely used to suppress rumor dissemination and control the spread of epidemics and diseases. However, achieving high accuracy and comprehensiveness in node influence ranking is time-consuming, and there are issues in using different measures on the same subject. The identification of influential nodes is very important for the maintenance of the entire network because they determine the stability and integrity of the entire network, which has strong practical application value in real life. Accordingly, a method based on local neighbor contribution (LNC) is proposed. LNC combines the influence of the nodes themselves with the contribution of the nearest and the next nearest neighbor nodes, thus further quantifying node influence in complex networks. LNC is applicable to networks of various scales, and its time complexity is considerably low. We evaluate the performance of LNC through extensive simulation experiments on seven real-world networks and two synthetic networks. We employ the SIR model to examine the spreading efficiency of each node and compare LNC with degree centrality, betweenness centrality, closeness centrality, eigenvector centrality, PageRank, Hyperlink-Induced Topic Search(HITS), ProfitLeader, Gravity and Weighted Formal Concept Analysis(WFCA). It is demonstrated that LNC ranks nodes effectively and outperforms several state-of-the-art algorithms.
Highlights
Complex networks are powerful methods for representing and studying the interactions among objects in the real world, it is an abstraction of complex systems
We propose a new method, called local neighbor contribution (LNC), for detecting influential nodes in complex networks of various scales
Intuitive and effective influential node detection: We consider the factors that affect the importance of nodes from two different perspectives, and node influence is measured through the combination of the influence contributions of the nodes themselves, the nearest neighbor nodes, and the nearest neighbor nodes
Summary
Complex networks are powerful methods for representing and studying the interactions among objects in the real world, it is an abstraction of complex systems. Intuitive and effective influential node detection: We consider the factors that affect the importance of nodes from two different perspectives, and node influence is measured through the combination of the influence contributions of the nodes themselves, the nearest neighbor nodes, and the nearest neighbor nodes This increases identification accuracy to some extent. Scalability: Compared with other methods, LNC can effectively and accurately identify influential nodes and greatly reduce computational cost (Fig. 2–Fig. 4, Table 4–Table 6) It is computationally simple and has low time complexity (the specific time complexity will be explained in Section III), it is suitable for large-scale networks.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
