基于多特征的热门微博预测算法研究

发布时间：2018-03-04 03:13

本文选题：微博舆情　切入点：微博预测　出处：《小型微型计算机系统》2017年03期 　论文类型：期刊论文

【摘要】：随着微博的迅猛发展,微博舆情已经成为研究热点.以新浪微博为研究对象,分析热门微博的影响因素,提出一种基于多特征的热门微博预测算法.首先,对微博的原始特征进行分析,从中提取关键特征.其次,利用信息增益算法,根据微博的传播特征对微博的热度进行度量.最后,结合BP神经网络算法,根据微博的内容和博主特征,预测微博的传播特征,并由此推算微博的热度来预测该微博能否成为热门微博.实验表明,该算法的查准率可以达到75%以上,F1度量值保持在78%左右,能够对刚发布的微博进行热度预测,适用于微博营销和舆情引导等领域.
[Abstract]:With the rapid development of Weibo, Weibo public opinion has become a hot research topic. Based on the analysis of influential factors of popular Weibo, this paper proposes a multi-feature based prediction algorithm for popular Weibo. First of all, This paper analyzes Weibo's original features and extracts the key features from them. Secondly, the heat intensity of Weibo is measured by using the information gain algorithm, according to the propagating characteristics of Weibo. Finally, the BP neural network algorithm is combined with the BP neural network algorithm. Based on Weibo's content and blogger's characteristics, we can predict the transmission characteristics of Weibo, and then calculate the heat of Weibo in order to predict whether it will become a hot Weibo. The experiment shows that, The precision of the algorithm can reach 75% or more and the F1 measure can be kept around 78%. It can predict the heat of Weibo which has just been published and can be used in the field of Weibo marketing and public opinion guidance.
【作者单位】：郑州大学信息工程学院;
【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(XMTGGCBJSZ05)资助河南省科技攻关项目(142102310531)资助郑州市科技攻关计划项目(141PPTGG368)资助
【分类号】：TP393.092;TP391.1;TP183

【参考文献】