Using the relationship of shared neighbors to find hierarchical overlapping communities for effective connectivity in IoT

Internet of things (IoT) aims at connecting everything and it would be a complex system with huge amount of resource. It can be abstracted as a complex network model. Discovering latent community structure will help understand large network structure, design high quality data integration, connection, communication mechanism based on suitable community topology in IoT. The diversity and complexity of real world make latent communities always be overlapped other than disjoint structure. Base on the observation that more common neighbors two nodes share, more chance they have to be in common community, in this paper we project a Neighbor-based Hierarchical Overlapped Community algorithm NHOC and a method for evaluating the discovered overlapping community quality. NHOC is polynomial and does not need community size or number in advance. It can discover overlapping or disjoint communities and with the parameters changing it can show different community structures in variable granularity. Experiments show that NHOC gives the more proper result than other algorithms when finding overlapping communities or disjoint communities.