Study On Node Propagation Importance For Multi-Dimensional Networks Based On Mutual Information

How to measure the importance of nodes in network topology is always a hot topic in the field of complex networks.Multidimensional problems in actual network,need to find suitable for multidimensional approach to the study of complex network node importance,and the node of strength is often a difference between,between nodes on the strength of the effect of each dimension is different,and as the change of time,the network nodes and between nodes strength is continually changing,therefore,the importance of the node degree is also in constant change.Aiming at the problems existing in the real network,this paper proposes a node importance updating method for multi-dimensional network information transmission.This paper first defines the auxiliary dimension to measure the overlap between dimensions and dimensions in multi-dimensional complex networks,calculates the average overlap rate between nodes and nodes according to the overlap ratio of each dimension,and proposes the definition of dimension density to improve the algorithm.The accuracy degree,considering the comprehensive influencing factors of the average overlap rate and the dimension intensity,the auxiliary filter approximate dimension value is obtained.Based on the research of the single-dimensional network mutual information method,the node mutual information calculation method applicable to the multi-dimensional network is proposed.Secondly,this paper gives the node shrinkage method for multi-dimensional network to calculate the change of mutual information,and uses the small world characteristics of complex networks to limit the shrinkage hops,so that the algorithm takes into account time efficiency and accuracy.Finally,based on the mutual information of the node itself and the mutual information change after the node shrinks,the degree of importance of the node based on the multi-dimensional network structure is comprehensively determined.Aiming at the problem of different interaction strength between nodes and nodes in real network,based on the multi-dimensional network structure,the definition of dimension traffic and the definition of dimension impact factor are proposed to measure the relationship between nodes and nodes in multi-dimensional networks with information dissemination.Dimension strength,further calculate the mutual information of each node,and give the node contraction method for multi-dimensional network of information propagation.Considering the multi-dimensional network based on information propagation,the degree of node importance is constantly changing.Partial update algorithm for node propagation importance.This article selects WikiElec large network and e-print arXiv super-large public data network to verify the effectiveness of the k jump to contract,and will be subject to the benchmark data compared with other algorithms,this algorithm was verified in selecting the appropriate shrinkage hop,its accuracy and efficiency than other algorithm has certain advantages.