Identifying Spread Blockers Using Overlapping Community Detection for Pandemic Management
Sajid Yousuf Bhat, Arjumand Akbar · 2025
A viral pandemic typically progresses through three stages, with the third stage marked by rapid community transmission at an exponential rate. At this stage, governments often resort to issuing lockdown advisories for the entire population to mitigate the spread. However, during the initial stages of the outbreak, precautionary measures or lockdowns can be targeted at a select group of individuals who hold the highest influence within the social network of the population, effectively curbing the disease&s;s spread by disrupting its transmission pathways. Furthermore, the same set of nodes can be used to plan effective priority-based vaccination strategies to further minimize the spread of the pandemic. Identifying influential nodes in social networks is a long-standing and challenging task in the field of complex networks research. One of the traditional approaches for identifying influential nodes in a social network is to use network node-centrality measures for ranking the nodes. However, computing global centrality measures is computationally exhaustive and often not scalable for very large-scale networks like the entire population of a country. In this chapter we present a novel study that empirically supports the concept that identification of influential spreaders in terms of betweenness centrality (referred to as spread blockers in this context) from a population can be alternatively modeled as identifying nodes in a network that has high community memberships, i.e., belong to multiple communities in the underlying overlapping community structure. The results are compared for various overlapping community detection methods and datasets and the best suitable candidate overlapping community detection methods for the desired purpose are reported.