A Study On The Influencing Factors Of Perceived Usefulness Of Online Reviews On E-commerce Websites

With the rapid development of information technology,internet technology and the economic society in various fields deeply mixing,e-commerce market booming,network shopping is penetrating into every aspect of life,becoming the most common way of consumption in the contemporary society.A large amount of online comment information has become a new way to spread product information in this network era.Consumers obtain real product information by reading online reviews information,eliminating information asymmetry caused by merchants describing and displaying products or services.At the same time,the emergence of the network water army,as well as the habitual reviews of consumers,and other factors have made the quality of massive online reviews intermingled,affecting potential consumers' access to information and misleading consumers' purchase decisions.The result also has a certain degree of impact on merchants and platforms.Therefore,identifying useful online comment information from a large number of comments has become the greatest demand of consumers.Although there are many researches on the usefulness of online reviews,due to the differences of research objects and methods,the research results are not uniform,and there are still defects.So it is very important to explore which factors will affect the perceived usefulness of online reviews.This research based on the related literature at home and abroad research,combinng Elaboration Likelihood Model and Hovland persuasion model as the theoretical basis,taking amazon website as an example,in the reviewers,the content of comments and the reviews reader to define 10 indicator which affecting the perceived useful of online reviews.It also establish the theoretical model of perceived usefulness of online reviews.This study selects the comments of electronic products of Amazon e-commerce website as the research sample,designs the crawler program based on the scrapy framework of python3.6,and combines Mysql and Redis database to collect the sample data needed for the research.On the basis of over-sampling preprocessing of these data,the proper data conversion method is adopted to quantify the indexes such as logarithmic conversion of comment length to improve their normality.The descriptive statistical analysis,Correlation analysis and multiplex collinearity test is carried out by using SPSS software.Finally,we use regression analysis to test the hypothesis of the model to verify the hypothesis resultWe get the following conclusion through empirical analysis,the length of reviews,the amount of information,the intensity of reviews emotion,the timeliness of reviews,the number of reviews star,the number of graphics,the number of reviewer' reply,the participation of reviewers,the authority of reviewers and the level of reviewers have a significant positive effect on the perceived usefulness of online reviews.According to the conclusions,this paper puts forward corresponding suggestions and countermeasures to consumers,merchants and e-commerce platforms.