Research on Wireless Network Consumer Satisfaction Prediction Method

With 5G network globalization, consumers have higher requirements for telecom operators' services. It is necessary to predict consumer satisfaction for analyzing consumer requirements. Based on the understanding of telecommunications services, the wireless network consumer satisfaction prediction is divided into three sub-predictive models: network quality, promotional activities, and tariff packages. At the same time, a hybrid sampling algorithm based on support vector machine (HS-SVM) which is used to classify the consumer satisfaction imbalance dataset is proposed to predict the consumer satisfaction of these three sub-predictive models, and the consumer's overall satisfaction is obtained by merging the results of the three sub-predictive models. The validity of the model is verified by wireless network consumer satisfaction dataset compared with the popular five separate classification algorithms and SMOTE combined with the five classification algorithms. The experimental results show that the F-value and G-mean of the proposed algorithm are improved. The proposed method has better classification performance and stronger robustness in the prediction of wireless network consumer satisfaction.