大数据中网络节点拓扑不稳定性的评估模型设计

发布时间：2018-05-05 15:25

本文选题：大数据 + 网络节点拓扑　；参考：《科学技术与工程》2017年15期

【摘要】：大数据中网络节点拓扑结构复杂且具有明显的不稳定性,当前网络节点拓扑不稳定性评估模型大多依据社会网络分析方法理论,从不同角度表示具有不同拓扑结构特性的网络元件,获取的评估结果不充分、不可靠。为此,提出一种新的大数据中网络节点拓扑不稳定性评估模型,通过节点收缩法对大数据中网络节点重要性进行评估。针对大数据中网络节点拓扑不稳定性的评估,结合风险函数丰富熵的内涵,将事件的风险函数在效用系数空间中的平均值看作网络的效用风险熵,依据得到的效用风险熵对效用风险熵权重进行计算。将节点收缩法和效用风险熵权重结合在一起,重新考虑权重的影响,给出新的大数据中网络凝聚度,得到大数据中网络节点拓扑不稳定性评估模型。实验结果表明,所提方法评估可靠性很高。
[Abstract]:In big data, the topological structure of network nodes is complex and has obvious instability. Most of the current evaluation models of network nodes' topological instability are based on the theory of social network analysis method. The network elements with different topology characteristics are represented from different angles, and the evaluation results are not sufficient and reliable. In this paper, a new evaluation model of topology instability of network nodes in big data is proposed, and the importance of network nodes in big data is evaluated by the method of node contraction. In view of the evaluation of topological instability of network nodes in big data and the rich entropy of risk function, the average value of event risk function in utility coefficient space is regarded as the utility risk entropy of network. The weight of utility risk entropy is calculated according to the obtained utility risk entropy. This paper combines the node contraction method with the utility risk entropy weight, reconsiders the influence of the weight, gives the network cohesion degree in big data, and obtains the topology instability evaluation model of the network node in big data. The experimental results show that the proposed method is highly reliable.
【作者单位】：安阳学院计算机系;
【基金】：陕西省教育厅科学研究项目(14JK2105) 西安外事学院教学改革研究项目(2015B08)资助
【分类号】：TP393.08

【相似文献】