在线评论有用性排序模型研究

发布时间：2018-06-22 08:41

本文选题：在线评论 + 有用性　；参考：《吉林大学》2017年硕士论文

【摘要】：根据中国互联网络信息中心(CNNIC)发布《2015年中国网络购物市场研究报告》显示,截至2015年12月,中国网络购物市场交易总次数达256亿次,年度人均交易次数62次。伴随着网络购物的蓬勃发展,购物信息巨增,研究表明,相比于商家展示的信息,消费者更愿意相信网络在线评论,也就是网络口碑。当前多数网络购物平台都拥有比较完善的售后评价系统,方便消费者对商品进行信息反馈,以供后续买家参考。但是伴随着电子商务的蓬勃发展,评价信息的数量呈现出指数级的增长,消费者在面对海量信息时,如何快速的识别可靠的信息,提高信息判断的准确性是一个亟待解决的问题。此种现象逐渐引起了国内外学者的关注,相关研究成果如雨后春笋,但目前学者的研究多数集中在评论识别、评论有用性影响因素等方面,对评论的排序过滤问题的关注度较少。针对目前的现状,本文从评论的有用性过滤排序入手,旨在将维度更丰富的评论信息优先呈现给消费者,提高消费者信息有用性感知。本文通过查阅以往文献,梳理在线评论有用性影响因素,构建本文的概念模型,通过定量研究方法对相关指标进行量化处理,并采用模糊层次分析法对三大指标进行权重的确定,并结合数据包络分析法对在线评论进行过滤排序,最终采用实证法进行验证。实验证明,本研究排序方法可靠,排序结果比较理想,能较好的处理当前在线评论大量冗余的问题,可明显提高消费者的在线评论有用性感知。对用户有效识别可靠信息,提高信息判断的准确性具有一定的理论与实践意义。
[Abstract]:According to the 2015 China online Shopping Market Research report released by the China Internet Network Information Center (CNNIC), the total number of transactions in China's online shopping market reached 25.6 billion by December 2015, with 62 transactions per person per year. With the rapid development of online shopping, shopping information has increased dramatically. Research shows that consumers are more willing to believe online reviews, that is, online word-of-mouth, than the information displayed by merchants. At present, most online shopping platforms have relatively perfect after-sale evaluation system, which is convenient for consumers to carry out information feedback for subsequent buyers. However, with the rapid development of electronic commerce, the quantity of evaluation information is increasing exponentially. How to identify reliable information quickly when consumers face mass information. It is an urgent problem to improve the accuracy of information judgment. This phenomenon has gradually attracted the attention of scholars at home and abroad, related research results such as springing up, but at present, most of the scholars focus on the identification of comments, comment usefulness factors, and so on. The problem of sorting and filtering comments is less concerned. In view of the present situation, this paper starts with the filtering and sorting of the usefulness of comments, aiming at giving priority to the more abundant comment information to consumers and improving consumers' perception of the usefulness of information. This paper reviews the previous literatures, combs the influencing factors of online review usefulness, constructs the conceptual model of this paper, and quantifies the relevant indicators by quantitative research methods. The fuzzy analytic hierarchy process (FAHP) is used to determine the weights of the three indexes, and the data envelopment analysis method is used to filter and sort the online comments. Finally, the empirical method is used to verify them. The experimental results show that the method is reliable and the sorting result is ideal. It can deal with the redundant problem of online reviews and improve consumers' perception of online comment usefulness. It is of theoretical and practical significance for users to identify reliable information and improve the accuracy of information judgment.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F724.6;F224

【参考文献】