集成复杂网络与多智能体仿真的人肉搜索效率研究  

Study of “Human Flesh Search” Efficiency Combining Complex Networks with Agent-based Model

作　　者：吴江[1,2] 贺超城 朱侯[3] 

机构地区：[1]武汉大学信息资源研究中心,武汉430072 [2]武汉大学电子商务研究与发展中心,武汉430072 [3]中山大学资讯管理学院,广州510006

出　　处：《情报学报》2018年第1期68-75,共8页Journal of the China Society for Scientific and Technical Information

基　　金：国家自然科学基金面上项目"创新2.0超网络中知识流动和群集交互的协同研究"(71373194)

摘　　要：为了探究人肉搜索中网民所处的网络拓扑及其性质对人肉搜索效率的影响,本文基于多智能体建模仿真的方法,从网民线下搜索、线上交流的微观层面构建人肉搜索仿真系统,将时间成本作为效率的评价指标,基于Python Network X实现了仿真系统。在平均度较小时,无标度网络下人肉搜索效率有明显优势,控制无标度网络的核心节点可以有效减缓人肉搜索效率,小世界网络重连概率在0.1左右人肉搜索效率最高;平均度较大时,小世界网络下人肉搜索效率有一定优势,单调的增加或减少平均度对人肉搜素效率的影响不明显,小世界网络重连概率的影响不明显,控制无标度网络核心节点的影响不明显。基于多智能体的人肉搜索仿真系统从微观层面对网民线下搜索、线上交流的过程进行模拟,得到网民所处网络结构对人肉搜索效率的影响规律,有助于政府和网络平台利用和控制人搜索现象,实现网络虚拟空间的和谐健康发展。This study explores the influence of network topology and its nature on the efficiency of human flesh search. Based on the multi-agent modeling and simulation method, the human flesh search and simulation system is constructed from the micro-level of offline users＇ search and online communication. The time cost is considered as the evaluation index of efficiency, and the system is realized based on Python NetworkX. When the average degree is small, the human flesh search efficiency of a free-scale network has obvious advantages. The core nodes of the free-scale network can effectively lower the efficiency. When the reconnection probability of a small-world network equals 0.1, the efficiency is the highest. When the average degree is large, the small-world network shows an advan- tage. Monotonically increasing or decreasing the average degree shows little influence on efficiency. The influence of the reconnection probability of the small-world network and controlling the core nodes of the free-scale network is not obvious. Based on multi-agent simulation, the human flesh search simulation system simulates the process of netizens＇ offline and online search from a micro level. It helps the government and the network platform to use and control the human flesh search phenomenon and achieve a harmonious development of virtual network space.

关 键 词：复杂网络 多智能体仿真 人肉搜索 搜索效率 用户行为 

分 类 号：O157.5[理学—数学]