基于水军信任惩罚的多维用户影响力度量模型
发布时间：2018-10-19 07:22
【摘要】：社交网络用户影响力度量是意见领袖识别的必要前提,然而目前的度量模型忽略了多维影响因素以及水军群体对度量模型真实性的影响.为此本文从网络结构、交互行为和交互信息三个维度来分析整合相关的影响因素,基于LeaderRank模型构建多维用户影响力度量-MUI模型,并在水军识别的基础上构建信任惩罚模型,以修正MUI模型建立MUISTP模型,实现社交网络用户影响力的真实性度量.以新浪微博平台大规模实际数据进行实验分析结果表明,与FBI,LeaderRank和Generative Graphical模型相比,MUI模型识别出的意见领袖更为准确,且可以实现更为有效的信息扩散.此外,经过水军信任惩罚后MUISTP模型较MUI能够实现更为有效的意见领袖识别.
[Abstract]:The measurement of social network users' influence is a necessary prerequisite for opinion leaders' identification. However, the current measurement model ignores the influence of multi-dimension factors and the influence of waterborne groups on the authenticity of the measurement model. Therefore, this paper analyzes the influence factors of integration from the three dimensions of network structure, interactive behavior and interactive information, constructs a multi-dimensional user impact metric MUI model based on LeaderRank model, and constructs a trust punishment model based on the recognition of naval forces. The MUISTP model is established by modifying the MUI model to measure the impact of social network users. The experimental results show that the opinion leaders identified by the MUI model are more accurate than the FBI,LeaderRank and Generative Graphical models, and can achieve more effective information diffusion. In addition, the MUISTP model can achieve more effective opinion leader recognition than MUI after the naval army trust punishment.
【作者单位】： 华南理工大学工商管理学院;
【基金】：国家自然科学基金(71571073) 广东省自然科学基金(2014A030313243)~~
【分类号】：TP393.0
本文编号：2280523
[Abstract]:The measurement of social network users' influence is a necessary prerequisite for opinion leaders' identification. However, the current measurement model ignores the influence of multi-dimension factors and the influence of waterborne groups on the authenticity of the measurement model. Therefore, this paper analyzes the influence factors of integration from the three dimensions of network structure, interactive behavior and interactive information, constructs a multi-dimensional user impact metric MUI model based on LeaderRank model, and constructs a trust punishment model based on the recognition of naval forces. The MUISTP model is established by modifying the MUI model to measure the impact of social network users. The experimental results show that the opinion leaders identified by the MUI model are more accurate than the FBI,LeaderRank and Generative Graphical models, and can achieve more effective information diffusion. In addition, the MUISTP model can achieve more effective opinion leader recognition than MUI after the naval army trust punishment.
【作者单位】： 华南理工大学工商管理学院;
【基金】：国家自然科学基金(71571073) 广东省自然科学基金(2014A030313243)~~
【分类号】：TP393.0
【相似文献】
相关期刊论文 前4条
1 吴振强;马建峰;;基于联合熵的多属性匿名度量模型[J];计算机研究与发展;2006年07期
2 刘学忠;刘增良;余达太;;基于AHP度量模型的安全管理度量方法[J];微计算机信息;2007年18期
3 吴江;徐俊刚;钟博;;一种安全监控事件风险评级度量模型[J];信息安全与通信保密;2009年12期
4 郑小蓉;;可信Web服务的度量模型[J];山东大学学报(理学版);2011年09期
相关硕士学位论文 前1条
1 曹颖;基于云环境的信任度量模型的研究与设计[D];华北电力大学;2015年
，本文编号：2280523
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2280523.html
