Modeling Of Complex Collaboration Network And Research On Information Dissemination In Network

The modeling of complex network is important to the research of complex network, theadvantages and disadvantages of model directly determine that if we can regenerate thestructure of real networks. In the paper, we propose a multi-parameter hybrid collaborationnetwork model (HCNM) using both similarity-strategy and copy-strategy. In this model, theprobability of two nodes linking is determined by the similarity of properties and thesimilarity of structures (Two nodes have many common friends) between them. Boththeoretical analysis and experiment can demonstrate that the HCNM model can accommodatedifferent real social networks. Only by adjusting the value of three parameters, we canregenerate the characteristics of social networks. Meanwhile, three types of models studiedabroad (the model based on similarity-strategy, the model based on copy-strategy and the purepreferential attachment model) can be seen as special cases of our HCNM model. Importantly,HCNM model has such characteristics which existing models have difficulty to fulfill all:(1)Not only the high average clustering coefficient, but also the high global clustering coefficientand obvious community structure (so many triangles);(2) New links can not only beconstructed between new nodes and old nodes, but also be constructed between old nodes.(3)The probability of new links which a node gains can not only be related linearly to its degree,but also be not linearly. In our paper, characteristics about complex network mainly involvesmall-world, scale-free, and clustering. The results of experiments, accommodating variouskinds of real networks, demonstrate that the HCNM model can regenerate characteristics ofreal complex networks，especially the high clustering. Importantly we can effectively modelamount of collaboration networks. For example, the regenerated collaboration network ofscientists can help us to analyze the special relationships between scientists, then to establishefficient relationship of scientific collaboration and promote scientific research.One of the most important objectives of studying complex network is to reveal thedynamic behavior of network, to research the flowing of information and to discuss theinfluence of static or dynamic network structure on information transfer. The paper proposes a kind of modified model, the Swapped Transfer Model (STM) of information in complexnetwork. The model formulates multiple aspects of information and introduces the growthfactor of information. It modeled the information transfer in complex network, based on‘swapping’ ideas. By the experiments of STM information transfer based on the small-worldmodel and HCNM model, we research special rules of information transfer and analyze themthoroughly. Experiments demonstrate information transfers quickly in the small-worldnetwork. The process of information transfer is influenced by different parameters. With thespecial experiments separating multiple parameters, we analyze outer and inner factors whichinfluence the process of information transfer.