基于内容和信任度的舆情扩散研究  

PUBLIC OPINION DIFFUSION PREDICTION ALGORITHM BASED ON CONTENT AND TRUST DEGREE

作　　者：陈振春 刘学军 李斌 

机构地区：[1]南京工业大学计算机科学与技术学院,江苏南京211816

出　　处：《计算机应用与软件》2017年第10期59-65,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(61203072);江苏省重点研发计划(社会发展)项目(BE2015697)

摘　　要：为提高微博网络中预测舆情转发规模和扩散深度的准确度,提出一种基于内容和信任度的舆情扩散预测算法。首先,依据微博网络中用户和舆情的内容信息,提取影响舆情扩散的特征指标,同时,结合用户间的信任关系,建立在单一邻居已转发舆情情况下用户转发行为的预测模型。继而,基于该模型和线性阈值模型,对多邻居已转发舆情的情况进行深入分析,最终完成对舆情转发规模和扩散深度的预测。实验结果表明,该算法显著提高了转发规模和扩散深度的预测准确性。To improve the public opinion＇s retweet scale and diffusion depth of prediction accuracy in microblogging network,a public opinion diffusion prediction algorithm based on content and trust degree is proposed. First,according to the content about users and public opinions from microblogging networks,the characteristic index influencing public opinion diffusion is extracted. Meanwhile,the algorithm obtains the trust relationship between users. Thus,the model of predicting user＇s retweeting behavior is established with only single neighbor who had retweeted the public opinion.Based on the above public opinion diffusion model and linear threshold model,a deep research on many neighbors who had retweeted the public opinion is done. Finally,the prediction about retweet scale and diffusion depth is completed.Meanwhile,the experimental results show that the algorithm improves the accuracy of predicting retweet scale and diffusion depth obviously.

关 键 词：舆情 内容 信任度 转发规模 扩散深度 预测 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]