Research On Ranking Important Nodes In Complex Networks Based On Improved H-index Algorith

Complex networks are a modeling tool that portray the interconnections among the various constituent elements of the complex system in the real world,and ranking of node importance is an important research topic in complex networks.Moreover,it also performs a crucial function in the structural evolution and stability of complex systems.Due to the interaction among individuals in the network has become increasingly complex and variable,while the real networks involve a large amount of temporal interactive information.In further,the temporal complex network constructed by introducing interaction information with temporal properties into the traditional static network can more explicitly represent the real world.Currently,among the existing methods for ranking the importance of nodes in static and temporal networks,most of them do not provide an accurate estimate of the nodes’ importance in complex networks,and have limitations in terms of resolution and accuracy.Taking into account the available research results,this thesis studies the node importance ranking algorithms in static networks and temporal networks respectively,and the principal work is summarized as follows.Regarding the problem of node importance ranking in static complex networks,this thesis improved the H-index algorithm and constructed a local nearest neighbor node H index important node ranking algorithm.Since the H-index algorithm assigns the same importance value to many nodes in the network,in this thesis,based on the analytical study of the H-index algorithm,comprehensively considering the degree information of the target node itself and their neighbors,the h-index values of the target node itself and their neighbors,and the different degrees of contribution provided by the neighboring nodes to the target node were combined.By the introduction of the secondary influence weighting coefficient of the nearest neighbor nodes to measure the indirect influence of neighbor nodes on the importance of the target nodes themselves,and thus proposed a local nearest neighbor node H(LNDH)index important node ranking algorithm,which solved the problem that most nodes have the same importance in the previous node importance ranking methods and could precisely distinguish the important nodes.The empirical results on the complex network dataset verified that the suggested algorithm could effectively assess the importance of nodes in different networks and had good differentiation ability.Regarding the ranking of important nodes in the temporal complex network,this thesis improved the LNDH algorithm and constructed the TLNDH index important node ranking algorithm based on the characteristics of the temporal network.Since the structural and functional properties of temporal networks evolve over time,the process of information propagation and diffusion naturally extends from the static network to the temporal complex network.Therefore,in this thesis,the temporal slicing network model was constructed based on the analysis of real temporal data.By improving the above node importance ranking algorithm in the static complex network,the LNDH algorithm was further applied to the temporal complex network,and then the TLNDH index important node ranking algorithm based on the temporal network was proposed.This algorithm comprehensively integrated the node themselves,their local attributes,and temporal information to jointly evaluate the importance of the temporal network nodes.By conducting experimental comparisons on the temporal network dataset,it is validated that the algorithm had good feasibility and availability for the ranking of critical nodes of the temporal network.