摘要
针对微博群落的识别与形成演化机理的获取问题,提出一种基于超图的微博群落感知方法.归纳微博用户的交互关系,提出以用户为节点、交互关系为边的微博用户交互关系超图模型;分析微博用户交互环境的情境特征,通过FP-TREE方法挖掘用户交互与情境特征的关联规则;根据关联规则对超图模型进行划分,得到具有相同情境的微博群落.以新浪微博为例进行了模拟验证,结果表明该方法能够感知导致微博群落形成的情境特征,且较传统数据挖掘方法能够更加准确地识别微博群落.
A method of micro-blog community awareness based on hypergraph is proposed to identify micro-blog community and explore its evolution mechanism.A hypergraph based relational model of micro-blog users is established by summarizing interactions among micro-blog users.Users are vertices,and interactions are edges in the hypergraph.Association rules among micro-blog interactions and context are mined using FP-TREE method by analyzing user profile and micro-blog message.Then,hypergraph is divided into communities with same context characters based on the association rules.A simulation test based on Sina micro-blog is performed to validate the effectiveness of the proposed method.The result shows that the method can be aware of the context of micro-blog community,and is more accurate than traditional data mining method in identifying the micro-blog communities.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2012年第11期120-126,共7页
Journal of Xi'an Jiaotong University
基金
国家自然科学基金资助项目(61033005)
关键词
微博
社交网络
微博群落
超图
情境感知
micro-blog
social networking
micro-blog community
hypergraph
context aware