文章导读
总览 评价 李瑞冰 , 崔鸿雁 * ( 北京邮电大学信息与通信工程学院,网络交换技术国家重点实验室,网络系统架构和融合重点实验室,北京,100876; ) 摘要: 研究人类行为分布对社会发展、城市建设、市场经济发展、信息传播等都有着极其重要的作用,因而受到
李瑞冰, 崔鸿雁*
(
北京邮电大学信息与通信工程学院,网络交换技术国家重点实验室,网络系统架构和融合重点实验室,北京,100876; )
摘要:
研究人类行为分布对社会发展、城市建设、市场经济发展、信息传播等都有着极其重要的作用,因而受到广泛关注。对于人类行为符合何种分布类型这一问题,现有的研究结果普遍表明,人类行为分布具有重尾特性对于使用何种函数能够更好地拟合人类行为分布的重尾特性这一问题,大多数学者认为幂律分布能够很好地对其进行拟合,但针对不同业务,也有学者提出Lognormal分布、两段分布等能更好地拟合人类行为分布。对于人类行为重尾特性是如何产生的这一问题,研究人员又提出了许多模型来解释。如经典的任务队列模型、基于兴趣的模型、基于记忆的模型等。对于人类行为时间方面的分布研究,主要集中在人类行为时间间隔分布的研究,因此本文综合分析了人类行为分布以及分布模型等问题。
关键词:
时间间隔分布;重尾特性;分布模型
LI Ruibing, CUI Hongyan*
(
1.Information and Communication Engineering School,Beijing University of Posts and Telecommunications,2.State Key Lab. of Networking and Switching Technology,3.Key Lab. of Network System Architecture Convergence,Beijing,100876; )
Abstract:
Research on human behavior distribution has attracted many attentions for its contribution to various areas like communication, medical treatment, financial, etc. For the study of human behavior distribution, there have been many conclusions. The existing results show that the distribution of human behavior has heavy tailed characteristics, which decline slower than the exponential distribution in the tail. For what kind of function can better fit the distribution of human behavior, most scholars believe that the power-law distribution can be suitable for fitting. However, some scholars propose the Lognormal distribution or two segment distribution can better fit it. Researchers have proposed many models to explain how the non Poisson properties of human behavior can be generated. Such as the queue model, the model based on human interest, model based on memory.The research on human behavior in time-field is mainly focused on the study of inter-event time distribution, so several typical human behavior distributions and models in time-field are introduced in the paper.
Tag:
点此返回栏目查看更多>>>参考论文