摘要
本文以新浪微博平台为数据采集平台,对微博信息传播的影响因素和效果进行数据分析,在借鉴信息传播四要素和流行三要素的基础上,总结出了影响微博信息传播的16个因素。首先通过对"风云人气榜"上随机抓取的320个新浪微博用户数据进行多元线性回归预测,实证得到粉丝数、工作时间和发布时间对微博信息传递有促进作用,而活跃度、休息时间和日期对信息传播有阻碍作用;然后利用爬取数据中提取的441 005个转发样本,通过逻辑回归、朴素贝叶斯和贝叶斯网络的概率模型分析,实证了社交类型对用户微博转发行为的影响最为显著,微博社交需求显著高于内容需求,并且根据ROC曲线得出综合类型对用户微博转发行为的预测最为精准。
In this paper,the influence factors and results of the data acquisition are analyzed based on Sina Weibo platform,summed up the 16 factors that affect the microblogging information dissemination,on the basis of the four elements and draw on the dissemination of information on the prevalence of the three elements. First,through the ＂Storm popularity list＂ of 320 randomly grab Weibo user data multiple linear regression forecasting,empirical get the number of fans,working time and release time on the microblogging messaging promote the role and activity,rest the time and date have hindered the spread of information. Then take the data extracted from the 441005 forwarding samples,through logistic regression,Naive Bayesian and Bias network probability model analysis,empirical social types of user microblogging forwarding behavior is the most significant,the social needs of micro blog is significantly higher than the content needs,according to the ROC curve to get the comprehensive type of user microblogging forwarding behavior prediction is the most accurate.
出处
《现代情报》
CSSCI
北大核心
2016年第3期22-26,共5页
Journal of Modern Information
基金
湖北省教育厅人文社会科学研究项目"基于用户需求的地方新闻网站内容生成创新研究"(项目编号:15G161)
关键词
新浪
微博信息
传播效果
回归分析
效果预测
影响因素
sina
micro blog information
dissemination effect
regression analysis
effect prediction
influence factors
作者简介
柯赟（1978-），女，副教授，研究方向：网络传播、网络技术、数据挖掘。
