基于任务队列的网络信息有效分类传播模型研究
Research on Effective Classification of Network Information Based on Task Queue
作者：杨林枫(重庆理工大学计算机科学与工程学院, 重庆 400054)；黄贤英(重庆理工大学计算机科学与工程学院, 重庆 400054)；刘小洋(重庆理工大学计算机科学与工程学院, 重庆 400054)；刘超(重庆理工大学计算机科学与工程学院, 重庆 400054)；刘万平(重庆理工大学计算机科学与工程学院, 重庆 400054)
Author：YANG Lin-Feng(College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China)；HUANG XianYing(College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China)；LIU XiaoYang(Chongqing University of Technology)；LIU Chao(College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China)；LIU Wan-Ping(College of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China)
收稿日期：2017-04-29 年卷（期）页码：2018,55(4):727-732
期刊名称：四川大学学报: 自然科学版
Journal Name：Journal of Sichuan University (Natural Science Edition)
关键字：非线性动力学；在线社交网络；网络传播模型；等待时间概率
Key words：nonlinear dynamics；online social networks；network propagation model；waiting time probability
基金项目：教育部人文社科青年基金 (16YJC860010,15YJC790061)； 重庆市教委任务社科一般项目(17SKG144)； 国家社科基金西部项目(17XXW004)； 教育部人文社会科学研究专项任务项目(16JDSZ2019)
中文摘要
针对在线社交网络信息传播模型在事件描述中没有对其利害分类、等待时间概率下降意义模糊，提出了一种非线性时变信息有效分类传播方法，并在此基础上建立了事件分类的E-C模型。首先利用动力学的网络传播模型、传播用户之间的社会网络关系与用户行为之间的联系，其次结合任务优先级、等待时间与概率发生函数的非线性时变关系分析了在线网络信息传播模型，最后引入N指数函数建立E-C模型。仿真结果表明：传播过程中等待时间概率图遵循幂律分布，改进后的模型对有利事件与传统模型作对比，在等待时间概率分布图中的效果有23.1%的提升，对于有害事件，则有21.8%的提升，理论仿真结果与真实数据的变化趋势一致，提出的E-C模型合理、有效。
英文摘要
Aiming at the problem that the online social network information propagation model is not classified in the event description and the latency of the waiting time is reduced, an effective method of non linear time varying information propagation is proposed.On this basis, the E C model of event classification is established.Firstly, the dynamic network propagation model is used to propagate the relationship between the social network relationship and the user behavior.Then, based on the nonlinear time varying relationship of task priority, waiting time and probability generating function, the model of online network information propagation is analyzed.And finally the N exponential function was introduced to establish the E C model.The simulation results show that the probability chart of waiting time obeys the power law distribution.The improved model compares the favorable event with the traditional model.The effect of the waiting time probability distribution is 23.1%, and 21.8% for the harmful event And the theoretical simulation results are consistent with the trend of real data.The proposed E C model is reasonable and effective.
【关闭】
