基于核心成员的社团结构发现算法研究

发布时间：2018-03-04 22:24

本文选题：核心节点　切入点：社团结构划分　出处：《山东师范大学》2017年硕士论文　论文类型：学位论文

【摘要】：复杂网络最近几年的发展趋势居高不下,复杂网络的研究在很多领域的应用都逐渐兴起,节点重要性的排序和社团结构划分是复杂网络的两个独立领域,但是他们之间有着一定的联系,核心的节点可能是小社团的核心,社团的核心成员也可能是网络中比较重要的节点。本文从以下三个方面进行研究:针对重要节点进行了研究。对传统的Ks分解算法和MDD算法进行了改进优化,提出了一种新的Ks值的算法,能更好的对节点进行排序。针对社团结构划分进行了研究。借助了万有引力公式提出了一种新的相似度指标运用到社团结构划分中,进而又提出了新的算法。新算法中把核心节点考虑进去,通过新排序算法得到的结果筛选合并出一定量的核心成员,进而再利用相似度进行社团结构划分。针对专利合作网络进行了研究。首先通过数据搜集整理分析,画出了网络图,并且分析了网络特征。其次构建了网络的模型,并且通过实验进行了模拟。最后把节点重要性算法和社团结构发现算法应用到了专利合作网络上。
[Abstract]:The development trend of complex network is high in recent years, the research of complex network is gradually rising in many fields, the ranking of node importance and the division of community structure are two independent fields of complex network. But there is some connection between them, the core nodes may be the core of the small community, The core members of the community may also be the more important nodes in the network. In this paper, the following three aspects are studied: the important nodes are studied, the traditional Ks decomposition algorithm and the MDD algorithm are improved and optimized. In this paper, a new algorithm of Ks value is proposed, which can sort the nodes better. The division of community structure is studied. A new similarity index is put forward in the division of community structure with the help of the formula of universal gravity. In the new algorithm, the core nodes are taken into account, and a certain number of core members are merged through the results of the new sorting algorithm. Then the similarity is used to divide the community structure. The patent cooperative network is studied. Firstly, the network diagram is drawn through data collection and analysis, and the network characteristics are analyzed. Secondly, the network model is constructed. Finally, the node importance algorithm and the community structure discovery algorithm are applied to the patent cooperative network.
【学位授予单位】：山东师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP301.6;O157.5

【参考文献】