Identifying Influential Spreaders in Complex Networks Based on Weighted Mixed Degree Decomposition Method

In any given real-world networks, there are only a few users, who potentially spread or control the information outbreak. Identifying these set of users is vital to prevent information outbreak or accelerate them. There are many legacy methods, like page rank, cluster rank and many centrality measure techniques to identify these influential spreaders. But researches have proved that they are not much effective to narrow down influential spreaders, and many new algorithms have been proposed to rank the influential spreaders. Most of the algorithms that were proposed calculate nodes influence purely based on networks structure. From the study, we have found that actual nodes influence not only depend on their connections but also depend on the relationship strength between them. So, we have come up with a new method named Weighted Mixed Degree Decomposition which considers the parameters like underlying relationship strength as its weight, exhausted degree, k-shell value etc., as input values and ranks top N influential nodes in a network. This proposed method is evaluated against simulated network and real-world datasets using of SIR epidemic model. The various experiment results show that considering relationship strength as a weight parameter can outperform classic methods and quantify the node influence more accurately.