摘要
针对微博用户行为问题,采集Sina微博实证数据分析了不同情境下的微博用户行为特征。分析结果表明:微博用户的发布行为存在周期性和阵发性,用户发布微博的时间间隔服从幂律分布,且幂指数与用户活跃性呈正相关关系,同时用户之间的交互能够进一步激发用户的发布行为。在此基础上,构建了兴趣、习惯、交互三重驱动模式下的用户行为动力学模型,通过数学推导可知,该模型能够获得幂指数为参数可调的幂律分布,且模型仿真结果与实证数据分析结果相吻合。
Aiming at the behavior of microblog users,Sina empirical data were collected to analyze microblog userbehavior characteristics in different circumstances. The results show that the user blog release behavior has periodic and outburst characteristics,the time interval of users releasing microblogs obeys a power-law distribution,and the relation between power exponent and user activity shows positive correlation. In the time between blog entries,the interaction among users can stimulate users to be more active microbloggers. On this basis,we construct a dynamic model driven by interest,habit,and interaction. Mathematical derivation shows that the model is able to obtain a power-law distribution with an adjustable power exponent. The simulation results of the dynamic model are consistent with the analysis results of empirical data.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2015年第9期1292-1296,共5页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(71271062)
中央高校基本科研业务费资助项目(HEUCFD1507)
关键词
时间间隔分布
兴趣
习惯
交互
行为动力学
微博
动力学模型
time interval distribution
interest
habit
interaction
human behavior dynamics
microblog
dynamic model
