有效量化指标的设计和标准化评价模型的建立,是传统公信力评价体系的两个重要问题。文章首先总结了当前国内外有关公信力评价体系的研究动态,并针对其中存在的问题做出相应述评,然后从社会计算的视角,提出基于在线社会媒体构建社会实体公信力评价体系的设计理念和研究框架。通过对包括用户评论数据和社会网络数据在内的大规模社会媒体内容挖掘,形成评价社会实体公信力的特征数据,并构造有效的评价模型。从数据、模型和服务3个层次上提出研究范式下的解决方案,为现有研究提供一个新的探索方向。
The traditional credibility evaluation system involves two important issues,the design of effective quantitative indexes and the building of standard evaluation models. This paper firstly summarizes the current domestic and foreign researches on the credibility evaluation system and reviews the existing problems. From the perspective of social computing,we propose a research framework of credibility evaluation for the social entities based on the internet media. Through the mining of large-scale social media content including user comments and social network data,this framework can extract the representation features for evaluating the credibility of social entities and build the effective evaluation models based on these features. This paper presents the technical solution from the three levels of data,models and services,and provides a new exploring direction for the existing researches.