改进PageRank算法挖掘社交网络关键用户  

Improved PageRank algorithm for mining key users in social network

作　　者：石立新 SHI Lixin(Information Center,Henan Provincial Administration for Market Regulation,Zhengzhou 450000,China)

机构地区：[1]河南省市场监督管理局信息中心,河南郑州450000

出　　处：《现代电子技术》2022年第12期95-99,共5页Modern Electronics Technique

摘　　要：为了准确而快速地挖掘社交网络中的隐藏关键用户,文中在分析经典PageRank算法平均分配权值缺点的基础上,为社交网络中的每个用户节点设置各自的权威度,并结合用户浏览网页的现实情况,模拟用户可以根据主观意向选择节点对应的链接操作,提出一种Au-2S-PageRank(Authority-2Step-PageRank)算法。该算法在程序设计上融合传统AuPageRank和2S-PageRank算法的优点,可解决权值分配和用户主观意向难以确定这两方面的问题。另外,使用推特数据集对Au-2S-PageRank算法、经典PageRank算法、MBUI-SFIM算法进行测试仿真。实验结果表明,相比另外两种数据挖掘算法,Au-2S-PageRank算法可以更加高效且准确地挖掘有向社交网络中的关键用户。On the basis of analysis on the drawback about weight average allocation of the classical PageRank algorithm,each user′s authority is set for his own node in the social network,the link operation corresponding to the node can be selected according to the subjective intention to simulate the user in combination with the actual situation of the web page browse of the user,and an Au-2S-PageRank(Authority-2Step-PageRank) algorithm is proposed to mine the hidden key users in social networks quickly and accurately. The algorithm can deal with the difficulties for determining the weight allocation and user ′ s subjective intention because it has fused the advantages of traditional Au-PageRank and 2S-PageRank algorithms in the program design. The Twitter dataset is used to conduct the testing simulation for Au-2S-PageRank algorithm,classical PageRank algorithm and MBUI-SFIM algorithm. The experimental results show that Au-2S-PageRank algorithm can mine key users in directed social networks more accurately and efficiently than other two data mining algorithms.

关 键 词：PAGERANK算法 社交网络 关键用户挖掘 权值分配 主观意向 仿真测试 

分 类 号：TN915-34[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]