摘要
微博用户节点重要性研究没有充分考虑到信息的高冗余性、传播速度快、时效性高等问题,为此构建基于贝叶斯模型的用户影响力计算方法。分析微博社会网络中用户的行为模式,使用贝叶斯网络模型进行用户节点属性的先验概率学习;通过人工标识重要用户节点,使用领域专家知识获取各属性的先验概率;对具有重要影响力的属性值进行学习,建立用户属性-影响力的贝叶斯网络模型,根据影响力排序得到微博社会网络用户节点重要性排序。实验结果表明,该方法可以显著识别重要用户节点。
Quantitative approaches for the rank of the importance of users nodes in Micro-blog social network merging concepts and findings from research on user micro-blogs＇ high redundancy,fast rate of transmission and high timeliness are missing.Based on the design science research paradigm,a Bayesian model based approach calculating user influence was proposed.The behavior pattern of users nodes and learns prior probability of user nodes＇ attributes were analyzed using a Bayesian network model.The important users nodes were manually identified,and the knowledge of expert fields was used to obtain the prior probability of each attribute.The attributes with a major impact were learnt,and the user attributes-influence Bayesian network model was established to get the order of importance of user nodes in Micro-blog social network according to the rank of influence of user nodes.To demonstrate its practical applicability,a dataset of Sina Micro-blog was used to evaluate the effectiveness of the proposed method in evaluating important users nodes.
出处
《计算机工程与设计》
北大核心
2016年第8期2050-2056,共7页
Computer Engineering and Design
基金
天津市科技型中小企业技术创新基金项目(12ZXCXGX33500)
关键词
重要性排序
用户节点
贝叶斯网络
微博
社会网络
舆情监督
rank of importance
user nodes
Bayesian model
micro-blog
social network
public opinion supervision
