基于在线评论的产品在线声誉研究

发布时间：2018-04-04 03:30

本文选题：在线评论　切入点：在线声誉　出处：《北京邮电大学》2015年硕士论文

【摘要】：声誉,自古以来,一直受到人们无以伦比的崇尚与坚持不懈的追求,任何人,拥有良好的声誉才能更好地立于天地之间,任何产品,拥有良好的声誉才能受到消费者的青睐。此外,在线评论正处于高速发展期,各大电商网站都给消费者提供了发表在线评论的平台,消费者购买相关产品后,可以自由地发表自己的购物体验以及产品使用心得,给后续的潜在消费者提供决策指导,由此发展而来的在线声誉也正受到越来越多的重视。本文以在线评论为基础,通过分词技术、情感语义分析构建数学模型,对产品的在线声誉进行研究。主要研究了基于在线评论的产品在线声誉测度方式,并探究了产品在线声誉对产品销量的影响作用。主要研究内容及成果如下： (1)以合作团队通过中文分词技术和情感语义分析技术获得的产品属性词库和评价情感词库为基础,构建产品在线声誉测度模型。在文献研究的基础上,本文认为在线声誉,是某一时间,人们基于自身对某人或某事物的多维度的感知,在互联网上通过用户创作内容(UGC)表达出来的对某人或某事物的特征或性质的评价。本研究摒弃了声誉传统的测度方式—心理态度量表法,转而借助文本分析和情感语义分析的结果,通过构建数学模型直接基于在线评论测度产品的在线声誉。 (2)在京东商城上选取了121款手机的相关数据,对本文所构建的产品在线声誉测度模型的可靠性进行了验证。为了证明在线声誉测度结果的可靠性,本文基于模型测度得道121款手机的在线声誉,然后先将其整体与权威第三方网站上提供的这些手机的综合排名进行相关性分析,相关系数为-0.320,再从中选取了两对互为竞品的手机进行对比分析,结果均验证了在线声誉测度模型是可靠的。 (3)基于121款手机的在线声誉,结合产品本身属性和在线评论数量特征构建计量经济学模型探究在线声誉对产品销量的影响。为排除变量间的多重共线性,首先进行了相关性分析,发现在线评论平均星级评分与在线声誉存在正相关关系,相关系数为0.493,因此,分别以在线声誉和平均星级评分为自变量与产品本身属性,即控制变量一起构建两个基础计量模型探究其对产品销量的影响,发现在线声誉对产品销量有显著正向影响,而平均星级评分对产品销量无显著影响；然后在基础模型的基础上逐一引入评论数量、差评率、在线声誉与差评率的交叉项,得到的研究结论有：在线声誉、评论数量对手机销量有显著正向影响,且在线声誉对产品销量的影响受到差评率的调节作用；差评率对产品销量有显著负向影响,这表明在研究在线声誉对产品销量的影响时,评论数量和差评率是不能忽视的。
[Abstract]:Reputation, since ancient times, has been unparalleled respect and persistent pursuit, anyone, have a good reputation can better stand between heaven and earth, any product, have a good reputation can be favored by consumers.In addition, online reviews are in a period of rapid development. All major ecommerce websites provide consumers with a platform for online reviews. After buying related products, consumers are free to publish their own shopping experience and experience in the use of products.The online reputation of future potential consumers is being paid more and more attention as a result of providing decision-making guidance.Based on online reviews, this paper constructs a mathematical model by participle technology and affective semantic analysis to study the online reputation of products.This paper mainly studies the online reputation measurement method of products based on online reviews, and probes into the effect of product online reputation on product sales.The main contents and results of the study are as follows:1) based on the product attribute lexicon and appraising emotion thesaurus obtained by cooperative team through Chinese word segmentation and affective semantic analysis, the online reputation measurement model of product is constructed.Based on the literature research, this paper argues that online reputation is based on people's multidimensional perception of someone or something at a certain time.An evaluation of the characteristics or nature of a person or thing expressed on the Internet through user-generated content.In this study, the traditional measurement of reputation, the psychological attitude scale, is abandoned, and the results of text analysis and emotional semantic analysis are used to construct a mathematical model to measure the online reputation of products directly based on online reviews.Based on the relevant data of 121 mobile phones in JingDong Mall, the reliability of the online reputation measurement model of products constructed in this paper is verified.In order to prove the reliability of the online reputation measurement results, this paper measures the online reputation of 121 mobile phones based on the model, and then analyzes the correlation between the overall and the comprehensive ranking of these mobile phones provided on authoritative third-party websites.The correlation coefficient is -0.320, and then two pairs of competing mobile phones are selected for comparative analysis. The results show that the online reputation measurement model is reliable.Based on the online reputation of 121 mobile phones, this paper constructs an econometric model to explore the impact of online reputation on product sales by combining the properties of products and the quantitative characteristics of online reviews.In order to exclude multiple co-linearity among variables, correlation analysis was carried out first. It was found that there was a positive correlation between the average star rating of online reviews and online reputation, and the correlation coefficient was 0.493.With online reputation and average star rating as independent variables and the attributes of the product itself, two basic econometric models are constructed together to explore the impact of online reputation on product sales, and it is found that online reputation has a significant positive impact on product sales.However, the average star rating has no significant effect on product sales, and then, based on the basic model, the intersections of comment quantity, bad rating rate, online reputation and bad rating rate are introduced one by one, and the results are as follows: online reputation.The number of reviews has a significant positive impact on mobile phone sales, and the impact of online reputation on product sales is regulated by the negative impact of the rating rate on product sales, which indicates that when studying the impact of online reputation on product sales,The number of comments and the poor rating rate cannot be ignored.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F724.6;F713.55

【参考文献】