大规模社交网络中高效的关键用户选取方法

发布时间：2018-09-17 16:11

【摘要】：针对大规模社交网络及其用户发布消息的历史数据,如何快速有效地选取具有较强信息传播能力的关键用户,提出了一种关键用户选取方法。首先,利用社交网络的结构信息,构建以用户为节点的有向图,利用用户发布消息的历史数据,基于Spark计算框架,定量计算由用户活跃度、转发交互度和信息量占比刻画的权重,从而构建社交网络的有向带权图模型;然后,借鉴PageRank算法,建立用户信息传播能力的度量机制,给出基于Spark的大规模社交网络中用户信息传播能力的计算方法;进而,给出基于Spark的d-距选取算法,通过多次迭代,使得所选取的不同关键用户的信息传播范围尽量少地重叠。建立在新浪微博数据上的实验结果表明,所提方法具有高效性、可行性和可扩展性,对于控制不良突发信息传播、社交网络舆情监控具有一定的支撑作用。
[Abstract]:Based on the historical data of large scale social networks and their users, this paper presents a method for selecting key users with strong ability to spread information quickly and effectively. Firstly, using the structure information of the social network, constructing the directed graph with the user as the node, using the historical data of the user to publish the message, and based on the Spark computing framework, the activity degree of the user is calculated quantitatively. In order to construct the directed weighted graph model of social network, we can use PageRank algorithm for reference to establish the measurement mechanism of users' ability to spread information. In this paper, the computing method of user's information transmission ability in large-scale social network based on Spark is presented, and then, the algorithm of d- distance selection based on Spark is given, which makes the information transmission range of different key users overlap as little as possible through multiple iterations. The experimental results based on Sina Weibo data show that the proposed method is efficient, feasible and extensible, which can support the control of bad information dissemination and the monitoring of social network public opinion.
【作者单位】：云南大学信息学院;
【基金】：国家自然科学基金资助项目(61472345,61562090) 云南省应用基础研究计划重点项目(2014FA023) 第二批“云岭学者”培养项目(C6153001) 云南大学青年英才培养计划项目(WX173602) 云南省教育厅科研基金资助项目(2016ZZX006,2016YJS005)~~
【分类号】：O157.5;TP393.09

【相似文献】