摘要
微博水军的存在极大地影响了微博上的信息质量。研究了微博环境下水军的特征,并从水军博主特征和博主所发微博特征这两个方面提取多个水军特征指标,使用熵值法确定各指标权重,并结合多指标综合指数法建立了微博水军自动识别模型。使用微博样本进行测试后,模型得到了82.3%的准确率和85.8%的召回率。
Micro-blog Water Army has highly affected the information quality. In this paper, We extract several characteristic indices from Micro-bloggers＇ features and features of Micro-bloggers＇ weibo by studying the characteristics of Micro-blog water army. Based on index weights of these indices defined by entropy method, this paper establishes an automatic recognition model of Micro-blog by using multi-index comprehensive index method. After testing, the precision ratio of this model reached 87. 3% and recall ratio reached 83. 6%.
出处
《情报杂志》
CSSCI
北大核心
2014年第7期176-179,共4页
Journal of Intelligence
基金
华东师范大学新型产业信息化实训中心"社交网络大数据的商业分析与数据挖掘平台建设"(编号:0092013011)项目资助
关键词
微博水军
综合指数
熵值法
特征选择
情感词表
micro-blog water army
comprehensive index
entropy method
feature selection
emotional vocabulary
