Hypernetwork Characteristics Analysis And Application Research

The emergence and development of complex network theory provides an efficient research method to study in characteristics of real-world network.Complex network that bases on general graph still has obstacles when it is applied to describe real network.For example,a general graph cannot represent complex relations of more than two nodes.Hypernetwork is a more comprehensive and effective approach to represent real-world than complex network,due to a hyperedge of the hypernetwork can contain more than two nodes.In this work,hypernetwork topology properties are analyzed by hypernetwork theory and hypergraph properties.We propose a modeling construction algorithm for classic QQ group hypernetwork and protein hypernetwork.And we analyze the topological indexes related to the empirical data,such as node hyperdegree,degree,average path length,clustering coefficient,sub-graph centrality,entropy and other parameters.Numerical analysis results of these topological indexes reveal that these two hypernetworks contain the following characteristics: both QQ group hypernetwork and protein hypernetwork have scale-free property;QQ group hypernetwork owns small-world characteristics;protein hypernetwork doesn't have small-world characteristics.In this work,we also recognize key nodes of these hypernetworks to explain some problems existing in real network.For example,the overflow of garbage information,the spread of network virus in social networks,and the identification of key proteins in protein networks.These negative effects or problems can be reduced or eliminated by recognizing these key nodes.