Data Analysis On Customer Satisfaction Based On Commodity Comment

With the rapid development of the Internet,different models of e-commerce companies in China’s rapid rise,the phenomennon of online shopping in China has been generally apparent,and even began to seriously affect the store business.After online shopping,the users would like to share their satisfaction and comments of the product online.The product reviews have a direct influence of the user’s choice of the product,they are valuable information,so mining information on the evaluation is extremely meaningful.However,the amount of comments is often large,the noise data is also much,so how to efficiently analyze the data of commodity reviews and how to extract the emotional of the text have become a key issue.This paper aims to build an emotion analysis model based on the data of Jingdong’s mobile phone reviews,compares the characteristic emotion of different mobile phones,constructs the satisfaction model,and calculates the customer satisfaction index.The comparison results provide reference information for the buyers and provide feedback to the manufacturer as well.And the satisfaction indexes can.be applied as valuable reference for Jingdong Mall personalized recommendation.This paper involves first introducing the background of the research,the status of the research,the significance of the research,the innovation of the subject,the framework of the research and the theory of related technology.Then do the sentiment analysis on the Jingdong comments corresponding to Huawei P9 and Apple iphone6 mobile phones,make comparison to see if any significant difference between them.By solving the problem,I would like to operate analysis in RHadoop’s multi-threaded and parallel environment.My approaches involves firsrt combining the R software and Hadoop software to crawling the mobile phone review data.Secondly,filtering out the noise data.Thirdly,segmenting the effective text data.Forthly,conduct Dependency Parsing to extract keywords in the text,such as feature words,modifiers,emotional words.Fifthly,constructing a comprehensive method of emotion analysis and adopting different calculation methods for different text structures.Sixthly,calculating the emotional value of the characteristic words.Finally,comparing the results of two mobile phones’ characteristic emotion values and customizing the satisfaction index system to build the user satisfaction model.The results show that Huawei P9 mobile’s competitive advangtage lies on phone camera,style,function,signal,acclaimed,meanwhile iphone6 mobile’s competitive advantage lies on quality,cost-effective,after-sale service.Huawei P9 mobile’s satisfaction index is slightly smaller than Apple iphone6.Moreover,I found out that although Huawei P9 mobile’s innovation is strong,its lack of high product quality and high-quality after-service still are the significant factors impacting sales.