Satisfaction Analysis of Manual Customer Service on E-commerce Platforms based on Machine Learning Models

With the rapid development of global digital transformation and artificial intelligence, the positioning and value of artificial customer service in customer service needs to be re-examined. This study focuses on customer satisfaction, explores the influence mechanism of manual customer service on it, and explores its key drivers in combination with quantitative analysis methods. Through the linear regression model, the significance of the influence of variables such as customer service response time, consulting question type, and communication channels were identified. In addition, the random forest model was used to analyze the characteristics and importance of each factor, and it was found that factors such as customer service response time and customer inquiry question had important impact on satisfaction, which further confirmed the calculation results of the linear regression model. This study provides a theoretical basis for enterprises to optimize the manual customer service training system and adjust humanized service strategies, and has practical value for improving customer satisfaction and improving the efficiency of manual customer service.