Research On Customer Satisfaction Of Fresh Food E-commerce Based On Text Mining

With the development of the Internet era,online shopping is so convenient that consumers are keen to purchase the required products online.In addition to online shopping for books,clothing,electronic products,etc.,fresh products have become popular products for online shopping.All kinds of fresh food e-commerce have emerged.However,due to the fact that fresh products are different from ordinary products and have characteristics such as perishability,timeliness,etc.With the improvement of the quality of life of residents,people have higher demands on the safety,quality,and convenience of purchasing foods.Fresh food e-commerce is still at the development exploration node.Satisfying the customer's needs,paying attention to the customer's consumption experience,and constantly making improvements and upgrades are what Fresh food e-commerce companies need to do in order to have more sustainable development,gain customer loyalty,and stand out in the e-commerce competition.This article uses the text mining method to research the customer satisfaction of the fresh food e-commerce.Firstly,three fresh food e-commerce with different operation modes and logistics modes were selected and Python was used to capture review texts.Analyze the text features of the captured comments.By using the word cloud image,the word frequency can be obtained,and the theme of the comment information can be displayed.Based on the analysis of network semantics,the evaluation words associated with high frequency words are obtained,and the advantages and disadvantages are analyzed;Then,TFIDF weight was used to construct the document entry matrix,and the feature attribute was used as the classification variable,and the whole sentence was used as predictive variable.Use CART algorithm to make classification decision trees to obtain the factors that each site needs to focus on improvement.Referring to the conclusion of the text feature analysis,the article extracts the characteristic emotional word pairs of the comments based on the part-of-speech template,applies the dictionary-based method to calculate the characteristic emotional values,summarizes the customer satisfaction evaluation index system,and calculates the index satisfaction with the corresponding weights.In the end,the author analyzes two kinds of text mining methods and conclusions based on feature analysis and feature emotion analysis,and analyzes the factors that influence customer satisfaction from the perspective of qualitative and quantitative analysis..Since Jd.com's and Motion optimization are preferably self-built logistics models,they have certain advantages in logistics distribution speed and delivery timeliness.,but Benlai has yet to be improved in the timeliness of logistics distribution and delivery service level.Summarize the advantages and disadvantages of each fresh food e-commerce business,and provide reasonable suggestions for the factors that need to be improved.