摘要
节点重要性评价对于网络大V的挖掘、网络病毒源的锁定和网络舆情的定向干预等方面具有重要应用前景。现有的节点重要性评价算法存在适应范围有限,评价结果与现实情况存在较大出入等问题。本文采用TOPSIS和灰色关联度相结合,设计了社交网络中节点重要性综合评价算法,提高了节点重要性评价算法的决策准确度和可适用于贫信息条件。最后利用小说《悲惨的世界》中人物关系网络数据集和Facebook一个月的数据验证本算法,并同TOPSIS算法和PageRank算法产生的结果比较,表明本算法更符合实际情况,并且具有较高的准确性。
The evaluation of node importance has important application prospects in the searching the very important people on net,the locking of network virus sources and the directional interference of network public opinion.Existing node importance evaluation algorithms have some problems such as limited scope of application,large discrepancies between the evaluation results and the actual situation.In this paper,TOPSIS and Gray Correlation are combined to design a comprehensive evaluation algorithm of node importance in social networks,which improves the decision accuracy of node importance evaluation algorithm and is applicable to the condition of poor information,and can well reflect the internal The difference between the trend of each attribute factor and the ideal solution.At last,the character relationship network data set in the novel"Les Miserables"and Facebook′s one-month data set are used to verify the algorithm.Compared with TOPSIS and PageRank,it shows that the algorithm is more realistic and with high accuracy.
作者
李晓龙
韩益亮
吴旭光
张德阳
LI Xiaolong;HAN Yiliang;WU Xuguang;ZHANG Deyang(Cryptographic Engineering College,Armed Police Engineering University,Xi′an,Shaanxi 710000,China)
出处
《燕山大学学报》
CAS
北大核心
2018年第5期444-450,共7页
Journal of Yanshan University
基金
国家自然科学基金资助项目(61572521)
军事科学研究计划课题基金资助项目(16QJ003-097)
