Research On Sentiment Analysis Basen On Online Comments Of Mobile Phone Users

In the face of fierce market competition,e-commerce platforms find that it is very important to think about products based on the level of consumers.When customers receive products from e-commerce platforms,they evaluate their satisfaction with the products based on their personal feelings of use,which fully reflects consumers’ feelings and opinions on the use of the products.Text reviews data from the consumer’s emotional point of view,not only can effectively provide enterprises with the improved new way of thinking,be helpful to product design optimization,and provide a detailed reference opinions and more effective marketing strategy,but also can provide effective suggestions to electric business platform,and make the platform to optimize customer experience,thus boosting the user activity and the improvement of loyalty.This paper crawls the comment information of users of Huawei,Xiaomi and other major brands and models of mobile phones,and preprocesses the comment text of users of Xiaomi,Apple and Huawei on JD platform to form a sentiment dictionary in the mobile field,and establishes a sentiment classification and analysis model based on machine learning rules.Secondly,based on the comment information of mobile phone users,further research on sentiment analysis and topic extraction is carried out.Sentiment analysis model is used based on different machine learning methods of SVM,NB and KNN.LDA model is formed based on users’ positive and negative comments,which can clearly understand users’ positive and negative experience of the product.The emotional concerns of users on different dimensions of the product are obtained and corresponding conclusions are drawn.Then,combined with the user comments of 14 Apple mobile phone products,the case analysis is carried out to draw the conclusion of sentiment analysis and user concerns of refined products.Finally,based on the conclusion of sentiment analysis,the paper puts forward relevant improvement suggestions for e-commerce platform and mobile phone manufacturers.