多维动态复杂网络中节点传播重要性算法研究

发布时间：2018-11-24 07:28

【摘要】：近年来,随着复杂系统的发展,越来越多的人开始使用互联网进行信息的共享和社交,网络的规模开始逐渐增大,网络结构也变得多样化。复杂网络可以看做是大量真实复杂系统的拓扑抽象,普遍具有小世界、无标度以及包含特定的社团结构等特性。随着复杂网络研究的深入,复杂网络的发展开始逐渐呈现出多维度共存的趋势。因此,研究复杂网络的多维度相关问题具有重要的学术意义和应用价值。复杂网络通常不是固定不变的,一般来说,复杂网络会随着时间的变化而发生边对的插入删除,故更好的处理复杂网络的动态变化成为了一个新的挑战。目前,现有的复杂网络相关研究大部分是针对单维静态复杂网络上进行,其研究成果难以适用于多维动态复杂网络场景。本文针对以上问题,在总结并分析现有的研究成果的基础上对多维度复杂网络的节点传播重要性进行研究。信息在多维网络中传播时,提出用维度相似性衡量维度间的差异,并考虑信息衰减对节点重要性的影响,进而给出节点传播重要性度量定义,以及该度量的静态计算方法。在此基础上,利用监测节点集监测网络的变化指导节点传播重要性度量的动态更新。本文通过在多个真实网络上运行节点传播重要性静态算法以及动态更新算法,实证了静态节点重要性算法针对于静态复杂网络节点重要性计算、动态节点重要性更新算法针对于动态复杂网络的社团更新均是高效准确的。
[Abstract]:In recent years, with the development of complex systems, more and more people begin to use the Internet to share information and socialize. Complex networks can be regarded as the topological abstractions of a large number of real complex systems, which generally have the characteristics of small world, scale-free and containing specific community structures. With the development of complex networks, the development of complex networks gradually presents the trend of multi-dimensional coexistence. Therefore, it has important academic significance and application value to study the multi-dimensional related problems of complex networks. Complex networks are usually not fixed. In general, the insertion and deletion of edge pairs will occur with the change of time, so it becomes a new challenge to better deal with the dynamic changes of complex networks. At present, most of the existing researches on complex networks are focused on single-dimensional static complex networks, and the research results are difficult to be applied to multi-dimensional dynamic complex network scenarios. In this paper, the importance of node propagation in multi-dimensional complex networks is studied on the basis of summarizing and analyzing the existing research results. When information propagates in multidimensional networks, the difference between dimensions is measured by dimensional similarity, and the influence of information attenuation on node importance is considered, and then the definition of node propagation importance measurement and the static calculation method of this metric are given. On this basis, the change of monitoring node set is used to guide the dynamic updating of node propagation importance measure. In this paper, the static algorithm of node propagation importance and the dynamic updating algorithm are run on many real networks, and the static node importance algorithm is applied to compute the node importance of static complex network. Dynamic node importance updating algorithms are efficient and accurate for community updating in dynamic complex networks.
【学位授予单位】：辽宁大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O157.5

【参考文献】