Customer Satisfaction Acquisition Method Of Online Reviews Based On Text Mining

In recent years,with the rapid expansion of the Internet and e-commerce and the rapid innovation of consumption patterns,online shopping has entered millions of households.As online shopping continues to heat up,the way customers express their opinions is constantly updated.Online reviews have become the most important feedback means for customers.Online reviews reflect the real evaluation of customers,which contains customers’ preference and satisfaction degree of products.Unlike traditional questionnaires,which are limited by time,region and cost,online reviews have the advantages of low cost,large quantity and unlimited expression.Therefore,more and more researchers and enterprise managers choose to mine satisfaction information from online reviews.Mining customer satisfaction from online reviews is the main concern of the paper.Specifically,there are three aspects.Firstly,the relevant researches on customer satisfaction are sorted out,the basic framework of mining satisfaction information from online reviews are summarized and the limitations of current satisfaction mining methods are analyzed.Secondly,a customer review satisfaction acquisition method based on Bert-Bi LSTM-CRF and Bert-Attention is proposed.In this method,Bert-Bi LSTM-CRF is used to extract feature words of reviews,Bert-Attention is used to conduct sentiment analysis of feature words,and customer satisfaction is mined based on feature words and sentiments.Finally,the proposed method is applied to practical problems,and experiments were conducted from real review data of Redmi mobile phones.The effectiveness is verified by comparing the quality of feature word extraction,sentiment analysis and satisfaction mining.The experimental results show that the satisfaction acquisition method of customer review based on BERT-Bi LSTM-CRF and BERT-Attention can effectively improve the quality of satisfaction mining and obtain more accurate satisfaction.For theoretical research,the basic framework of customer satisfaction mining based on online reviews is summarized,which can provide an effective procedure of satisfaction mining.Then the satisfaction acquisition method of customer review based on BERT-Bi LSTM-CRF and BERT-Attention is proposed,which effectively improves the quality of satisfaction mining and provides an effective data-driven approach for subsequent research on customer satisfaction.For practical application,customer satisfaction is an important way for enterprises to obtain customer feedback,which plays guiding role in the development of enterprises.The proposed satisfaction acquisition method can provide more accurate satisfaction information for enterprises,so as to help enterprise decision-makers formulate improvement methods in service quality,production technology and marketing strategy better and faster.Further,the service quality,product quality and marketing ability of enterprises are improved.Finally,the aim of improving the economic benefits of enterprises is achieved.