The Impact Of Online Comments Of E-Commerce Platform On Sales Of Large Household Appliances From The Perspective Of Text Mining

With the development of e-commerce,online shopping has become one of the main forms of consumer consumption.Online shopping can break through the constraints of time and space,making shopping more convenient.But the network commodity transaction is virtual and the description of commodity is abstract,which makes consumers unable to fully understand the quality information and other information of commodity.Online comment is an important channel for consumers to establish their perception of commodities,and it is one of the main factors affecting consumers'buying behavior.As a typical durable consumer goods,large household appliances have the characteristics of longer service life,fewer repeated purchases and higher prices.Therefore,compared with fast moving consumer goods,the consumption decision-making process is more complex and the influencing factors are more diverse.Therefore,it is of great significance to study the impact of online reviews on the sales of refrigerators,a large household appliances.This paper takes refrigerator as a research sample,and then grabs the comment data according to "default ranking" to accurately reflect the real shopping process of consumers,and uses text processing technology to calculate the emotional value of the comment instead of the emotional polarity to improve the credibility of the empirical analysis conclusion.Then,the first model studies whether supplementary comment have an impact on refrigerator sales,whether supplementary comment have a greater impact on refrigerator sales than one-time comments on refrigerator sales,and whether quality-related comments and service-related comments have an impact on refrigerator sales.Finally,in order to further study the impact of quality-related reviews on refrigerator sales,LDA is used to classify the topics of quality-related reviews on the basis of model one experimental conclusion.According to the results of thematic classification,the three dimensions of refrigerator appearance.refrigeration insurance effect and noise are defined as the perspective of quality-related reviews and model 2 is proposed.Secondly,the SVM algorithm is used to classify the quality reviews,and the impact of emotional value of three-dimensional reviews on refrigerator sales is studied.Finally,according to the results of the test,some suggestions are put forward for the refrigeratory manufacturer's production and the online retailer's marketing strategy.