摘要
随着在线社交网络的快速发展,越来越多的人开始利用微博等交流工具来传播信息或分享观点.微博已经成为一个表达公共舆论的重要传播媒体,研究微博中的信息转发规律对于热点挖掘、舆情监控、品牌营销等有着重要意义.以新浪微博为例,从发布用户、接受用户、微博内容三个方面进行特征提取,结合支持向量机SVM分类器进行用户去重、垃圾用户滤除,将提取的特征输入到预测算法中,建立逻辑回归模型,实现对微博转发预测.与传统同类预测模型进行对比试验,验证本文方法的正确性与有效性.
With the rapid development of online social networks, more and more people begin to use communication tools such as mi- croblog to spread information or share ideas. The microblog has become an important media to express public opinion,the research on the rules of the information in the microblog retweeting has the important significance for hot spot mining, public opinion supervision, brand marketing. A case study of sina microblog, we carry on the feature extraction from three aspects including the released user, accept user,microblog content,combine with support vector machine SVM classifier for removing user repetition, filtering spam,estab- lish logistic regression model, input the characteristics extracted to the prediction model, and complete the prediction of microblog. Compared with the traditional similar model test and we verify the correctness and effectiveness of this method.
出处
《小型微型计算机系统》
CSCD
北大核心
2016年第8期1651-1655,共5页
Journal of Chinese Computer Systems
基金
郑州大学新媒体公共传播学科招标课题阶段性成果项目(XMTGGCBJSZ11)资助
河南省科技攻关项目(142102310531)资助
关键词
微博
特征提取
SVM分类器
逻辑回归模型
转发预测
microblog
feature extraction
SVM classifier
logistic regression model
retweet prediction