Abstract:Networks with a community (or cluster) structure underlie many social and biological phenomena. In such a network, individuals tend to form sparsely linked local communities, each having dense internal connections. This dissertation explores the different dynamics of complex clustered networks, revealing a new set of rules that show how the dynamic properties are affected by the clustered structures. The dynamics of information propagation on clustered networks is studied by using a three-state epidemic model with a unit spreading rate. A resonance-like phenomenon is uncovered: the information lifetime on the network can be maximized by the number of clusters. Synchronization in complex, clustered random networks is found to be determined by the interplay between inter-cluster and infra-cluster links The network is most synchronizable when the numbers of the two types of links are approximately equal. In the presence of a mismatch, increasing the number of infra-cluster links, while making the network distance smaller, can suppress or even destroy the synchronization. For clustered networks with regular subnetworks, as the density of intra-cluster links is increased, the network exhibits strong and weak synchronizability in an alternating manner. A theory based on analyzing the eigen-values and eigenvectors of the coupling matrix is provided to explain this phenomenon. For gradient clustered networks, the synchronizability can be optimized by the strength of the gradient field. A remarkable finding is that, if the gradient field is sufficiently strong, synchronizability of the network is mainly determined by the properties of the subnetworks in the two largest clusters. A model of cascading in complex clustered networks based on physical analysis and numerical computations is developed for the key ingredients of traffic dynamics in typical clustered networks. An effective strategy is proposed for preventing cascading breakdown. By focusing on network synchronizability, it is found that globally coupled networks and random networks are scalable, but locally coupled regular networks are not. Scale-free networks are scalable for certain types of node dynamics. For a typical clustered network, as its size is increased, the synchronizability can be maintained or even enhanced but at the expense of deterioration of the clustered characteristics.

Evolving the Attribute Flow for Dynamical Clustering in Signed Networks

CoHomo: A Cluster-Attribute Correlation Aware Graph Clustering Framework

Information spreading with relative attributes on signed networks

Weighted Flow Diffusion for Local Graph Clustering with Node Attributes: an Algorithm and Statistical Guarantees

Clustering Large Attributed Information Networks: an Efficient Incremental Computing Approach.

A dynamic evolutionary clustering perspective: Community detection in signed networks by reconstructing neighbor sets

Optimization of Graph Clustering Inspired by Dynamic Belief Systems

Evolving Scale-Free Network Model With Tunable Clustering And Apl

Representative Community Detection Algorithms for Attribute Networks

Evolutionary Dynamics Analysis of Complex Network with Fusion Nodes and Overlap Edges

Overlapping graph clustering in attributed networks via generalized cluster potential game

Collective Adaptation Evolution Of Weighted Complex Networks: On Syncronizability Dependence

Dynamics and Security of Complex Clustered Network Systems

Accurate Detection of Hierarchical Communities in Complex Networks Based on Nonlinear Dynamical Evolution.

Partition Signed Social Networks Via Clustering Dynamics

Evolving K-Graph: Modeling Hybrid Interactions in Networks.

Multi-Order Clustering on Dynamic Networks: on Error Accumulation and Its Elimination

Modeling, Analysis and Validation of Evolving Networks with Hybrid Interactions

A Speedup Spatial Rearrangement Algorithm for Dynamic Network Visualization.

Node Attribute-enhanced Community Detection in Complex Networks

Identifying Influential Nodes in Large-Scale Directed Networks: The Role of Clustering