微博信息可信度评估的数据起源方法  

Evaluation of Weibo credibility based on data provenance

作　　者：张子良 董红斌[2] 谭成予[1] 梁意文[1] Zhang Ziliang;Dong Hongbin;Tan Chengyu;Liang Yiwen(School of Computer Science,Wuhan University,Wuhan 430079,China;International School of Software,Wuhan University,Wuhan 430079,China)

机构地区：[1]武汉大学计算机学院,武汉430079 [2]武汉大学国际软件学院,武汉430079

出　　处：《计算机应用研究》2018年第11期3330-3334,共5页Application Research of Computers

基　　金：国家自然科学基金面上项目(61170306)

摘　　要：针对现有微博可信度评估中缺少对传播信息的分析,提出一种基于数据起源的微博可信度评估方法,对微博的信息来源和传播过程进行研究。以新浪微博为研究对象,分析用户的关键指标以及微博信息的传播场景,建立微博的PROV数据起源模型,记录微博信息的来源和传播过程的数据起源信息,并验证其有效性。结合传播过程中用户的可信度,利用数据起源信息,完善微博可信度评估的参考依据,弥补了现有一般评估方法中传播信息缺失的不足。Aiming at the lack of analysis of transmission information in microblog credibility evaluation,this paper proposed a method based on data provenance to research the information source and transmission process.Taking Sina Weibo as the research object,this paper analyzed the key benchmarks of users and the scene of information transmission,built a PROV-model to describe Weibo,recorded the provenance information in transmission,and verified the validation of the provenance information.By combining the users’credibility in transmission trace and using data provenance information,this paper optimized the reference of Weibo credibility,made up for transmission information deficiency in existing general evaluation methods.

关 键 词：微博 可信度 数据起源 PROV模型 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]