Cigarette retail customer comprehensive satisfaction evaluation model based on multi-stream deep residual GRU network
In order to comprehensively and accurately grasp the satisfaction level of cigarette retail customers with each district and county branch in the tobacco industry in cigarette marketing work,continuously improve customer service methods and continuously enhance customer service,we proposed a cigarette retail customer comprehensive satisfaction evaluation model based on multi-stream deep residual GRU neural network.The model was based on 2 300 000 retail customer satisfaction questionnaire data during 2015-2022 for learning,classifying customer satis-faction with services into six levels and conducting corresponding sentiment analysis to get the final sentiment classi-fication results as the basis for comprehensive satisfaction evaluation.Simulation comparison experiments and actual data analysis results showed that this model had strong text feature extraction ability and could obtain better senti-ment classification results compared with traditional customer comprehensive satisfaction evaluation methods.It also provided new perspectives and tools for the management of retail markets in related industries.
deep learningdeep residual networkGRU networkcigarette retailcustomer comprehensive satis-faction evaluation
万方数据
维普
