User Comment Analysis System Design And Implementation Based On Text Sentiment Analysis Technique

The continuous development of Internet technology to promote the development of e-commerce,online shopping and people's lives more and more closely,in this case will produce many products and services for a variety of user comments data,these comments on the user and business are concerned Has a very important use value.But the Internet information changes very quickly,the network user comments on the amount of information is large,but also has a strong arbitrariness,which led to a lot of information in a lot of spam.If the manual way to deal with will consume a lot of manpower and material costs,the effect is not satisfactory.So,the need for computers to help users and businesses to obtain information and analysis,text emotion analysis technology can deal with this task well.Based on the theory of emotional analysis of text,this paper designs a set of user feedback emotional analysis system based on the emotional analysis of product reviews.The system can automatically obtain the information on the Internet according to the needs of users,Information is used to filter,and then use the emotional analysis technology to analyze and deal with these comments data,automatically identify the user's product and product attributes of the degree of preference,summed up the positive and negative information about the product and the product attributes involved.Finally passed.Visualization technology will show the results to users and enterprises the main research contents are as follows:(1)Based on the two-way propagation algorithm,a product attribute and user view extraction method based on rule and statistic are proposed.According to the particularity of Chinese language words,this kind of rule and the two-way communication in Chinese language are combined with the statistical algorithm,which can improve the accuracy of extracting product attributes and user opinion words,and then through the point mutual information Method to extract the information to filter,the redundant information removed,and finally the formation of product attributes and the corresponding relationship between the user point of view;(2)Design an improved dictionary method,introduce the support vector machine algorithm in machine learning,use the support vector machine to classify the text,and then produce a set of emotional dictionaries,combined with the previous classification results through the emotional dictionary to complete the Comments on the emotional polarity of the identification work,and ultimately achieved very good results.(3)Design a set of visual display scheme,the text of the emotional analysis of the results through the information visualization of the way out,according to the different results of different graphics to display.(4)Finally,the effect of the whole system is evaluated,and the satisfaction of the user experience is investigated with the effect of visual design.Finally,the results are analyzed and summarized.