基于信息密度的微博突发话题检测模型研究  

Research on the Microblog Bursty Topic Detection Model Based on Information Density

作　　者：王征[1,2] 王林森[1,2] 赵磊[1,2] 

机构地区：[1]西南财经大学经济信息工程学院,四川成都611130 [2]西南财经大学研究生院,四川成都611130

出　　处：《情报理论与实践》2016年第3期125-129,共5页Information Studies:Theory & Application

基　　金：国家社会科学基金青年项目"群体性事件管理推演与应对措施验证研究"的成果;项目编号:14CGL050

摘　　要：在分析当前微博突发话题检测模型存在的问题的基础上,提出了基于信息密度的新型检测模型MBID。该模型通过动态滑动窗口采集微博信息流,以话题树进行信息归纳,并最终通过窗口和话题树的信息密度变化进行监测,发现突发话题,并从主题树中抽取描述词。仿真实验表明,MBID模型较之其他两种模型,具有较高的检测精度、全面性和较高的处理效能。Based on the analysis of the existing problems of the present microblog bursty topic detection model, this paper puts forward a novel detection model based on information density. The MBID model uses dynamic sliding windows to get microblog infor- mation flows and summarizes information through topic trees. Finally, the paper detects and discovers topic through the information density of windows and topic trees, and extracts the description words from topic trees as well. The simulation results show that MBID has better detection accuracy, coverage and processing performance than the other two traditional models.

关 键 词：突发话题 微博 信息密度 检测 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术] G206[自动化与计算机技术—计算机科学与技术]