新浪微博信息传播路径阻碍因素分析及传播效果预测  

Sina Micro-blog Information Transmission Path Obstacle Factor Analysis and Communication-effect Prediction

作　　者：田向国[1] 肖林鹏[1] 刘铁英[2] 张小莉[2] 

机构地区：[1]天津体育学院体育教育训练三系,天津300381 [2]河北大学计算中心,河北保定071002

出　　处：《情报科学》2016年第5期91-94,共4页Information Science

基　　金：天津市高等学校人文社会科学研究项目(2011FDY30)

摘　　要：建立了基于多元线性回归的传播效果宏观预测模型及基于概率模型的传播效果微观预测模型,对随机抓取的300个新浪微博用户数据进行多元线性回归预测,证实粉丝数对用户微博信息转发数量有着显著的影响。对抓取数据中提取的439607个转发样本,进行基于逻辑回归、朴素贝叶斯和贝叶斯网络的概率模型分析,证实社交类型对用户微博转发行为的影响最为显著。通过模型的实证得到了影响新浪微博信息传播的因素,对微博信息的传播效果有一定的预测价值。This paper establishes the macroscopic prediction model based on multivariate linear regression and the micro-cosmic prediction model based on probability model of communication effect. By random grabbing 300 micro-blogging us-er data,the author conducts a multiple linear regression prediction,proves that the number of fans is a significant factor af-fecting the number of micro-blogging forwarding information. The paper analyzes the 439,607 transponders samples basedon Logistic regression, Naive Bayes and Bayesian Networks. Results show that social type has the most significant influenceto the user micro-blogging forwarding behavior. By empirical evidence we obtain the influence factors of sina micro-bloginformation communication. The research has predictive value.

关 键 词：新浪微博 传播效果 回归分析 概率模型 ROC曲线 

分 类 号：G206.3[文化科学—传播学]