摘要
本研究采用实验法,利用三个可衡量关注度的变量设计规则算法,以每天和每小时作为检测周期,对逸仙时空BBS中的主题帖进行筛选并预警,达到实时监控网络舆情突发异常的效果。
In this experimental study, three measurable variables of concern are used for designing the rules and algo- rithm. Choosing per day and per hour as the test cycle respectively, the topic posts of Yat-sen Time BBS were filtered and warned in time, achieving the effect of real-time monitoring online public opinion unexpected exceptions.
出处
《图书情报知识》
CSSCI
北大核心
2012年第1期43-51,共9页
Documentation,Information & Knowledge
基金
广东省教育厅产学研合作专项资金项目"网络舆情智能监测与分析系统"(2007A090302027)成果之一
关键词
网络舆情
突发异常
模式识别
关键算法
Online public opinion Unexpected exception Pattern identification Key algorithms