计算机工程与应用 ›› 2020, Vol. 56 ›› Issue (24): 123-129.DOI: 10.3778/j.issn.1002-8331.2005-0040

• 模式识别与人工智能 • 上一篇    下一篇

节点传播能力的偏好随机行走的信息传播方法

李维勇,孔枫,张伟,陈云芳   

  1. 1.南京信息职业技术学院 网络与通信学院,南京 210023
    2.南京邮电大学 计算机学院,南京 210023
  • 出版日期:2020-12-15 发布日期:2020-12-15

Node-Diffusing Capability of Biased Random Walks of Information Diffusion Method

LI Weiyong, KONG Feng, ZHANG Wei, CHEN Yunfang   

  1. 1.School of Network and Communication, Nanjing Vocational College of Information Technology, Nanjing 210023, China
    2.School of Computer Science, Nanjing University of Posts and Telecommunications, Nanjing 210023, China
  • Online:2020-12-15 Published:2020-12-15

摘要:

随机行走是社交和生物系统中用来模拟传播过程的标准化工具,针对真实社交网络中任意程度的有偏随机行走过程和由优先转移概率定义的偏向性,提出了一种新的用于研究社交网络的影响力传播范围最大化的方法,称之为基于节点传播能力的偏向性随机行走的网络信息传播方法(DCID),该方法随机从网络中选择一个信息传播源节点,使得该模型更加符合真实的社交网络;通过节点能承受的传播信息的内容量参数以及偏向性随机行走的参数来作为节点的优先转移概率;并通过影响力传播函数来衡量信息的影响力传播范围,以此达到信息传播范围的最大化。从真实的不同规模的社交网络中选定这两个参数值,并验证了提出的模型在不同规模社交网络中信息的覆盖率和算法运行时间的性能上有所提升。

关键词: 信息传播, 偏向性随机行走, 节点传播能力, 社交网络

Abstract:

Random walks are a standard tool for modeling the spreading process in social and biological systems. Aiming at any level of biased random walks and the bias defined by the priority transition probability in real social networks, a new method is proposed for studying the influence spread of social networks to maximize the range of spread, which is called information diffusion based on node-diffusing capability of biased random walks in social network(DCID). This method randomly selects an source node of information diffusion in the network and makes the model more in line with the true social networks. By the parameters of the node can be tolerated the amount of information and the parameters of the biased random walks are used as the node’s priority transfer probability. And by the influence diffusion function to be the measured the influence spread range of the information so as to maximize the information spread range. The values of these two parameters are selected from the real social networks of different scales, and it is verified that the proposed model has improved the coverage of information diffusion and the performance of running time in social networks of different scales.

Key words: information diffusion, biased random walk, node-diffusing capability, social network