计算机科学与探索 ›› 2018, Vol. 12 ›› Issue (5): 777-784.DOI: 10.3778/j.issn.1673-9418.1705028

• 人工智能与模式识别 • 上一篇    下一篇

基于手机通话网络的人类交互行为分析

李  凯1,2,张锡哲3+,申毓佩3,陈恩红1,2   

  1. 1. 中国科学技术大学 计算机科学与技术学院,合肥 230022
    2. 大数据分析与应用安徽省重点实验室,合肥 230022
    3. 东北大学 计算机科学与工程学院,沈阳 110819
  • 出版日期:2018-05-01 发布日期:2018-05-07

Analysis of Human Interactive Behavior Based on Phone Communication Networks

LI Kai1,2, ZHANG Xizhe3+, SHEN Yupei3, CHEN Enhong1,2   

  1. 1. School of Computer Science and Technology, University of Science and Technology of China, Hefei 230022, China
    2. Anhui Province Key Laboratory of Big Data Analysis and Application, Hefei 230022, China
    3. School of Computer Science and Engineering, Northeastern University, Shenyang 110819, China
  • Online:2018-05-01 Published:2018-05-07

摘要: 移动通信行为是社会生活的重要组成部分,手机通话网络是一种重要的社交网络,深入分析手机通信网络的拓扑特征对认识社交网络和人类行为特性具有重要的意义。基于复杂网络相关理论,提出了多重时变通信网络模型,在此基础上,对一个手机通信数据集进行分析挖掘,得到了人类动力学角度的行为模式。首先构建了手机通话网络,然后从群体和个体两个尺度对网络度分布、度演化、通话时长、呼叫间隔等指标进行了研究,发现网络整体上呈现明显的无标度特征,个体行为与节点度密切相关,而群体度演化过程则显示了特定地区社会生活特征,相关结果对于社会网络分析和人类行为动力学研究具有重要的意义。

关键词: 人类动力学, 手机通信, 特征挖掘, 复杂网络, 拓扑分析

Abstract: Mobile communication plays an important role in social life, thus the mobile phone network is an important social network. Effective analysis of the topology characteristics of mobile network is important to understand the social network and human behaviors.This paper proposes a multiple time-varying communication network model based on network theory, and obtains the human dynamics behavior pattern by investigating a mobile phone communication dataset. This paper first builds the mobile phone network, and then analyzes?the degree distribution, degree evolution, call time and call interval from both group and individual scales. The results show that the whole network has obvious scale-free characteristics, and the individual behavior is closely related to its degree, while the degree evolution of group shows the characteristics of social life in a specific region. The results are of great significance for social network analysis and research of the human dynamics.

Key words:  human dynamics, phone communication, characteristics mining, complex network, topology analysis