Research And Application Of Express Customer Service Performance Evaluation Based On Improved Decision Tree Algorithm

Our courier industry has experienced nearly thirty years of development,thanks to the rapid development of online e-commerce in recent years,the domestic express delivery industry even based on annual growth rate of 30 per cent of development and growth,with the express demand for logistics services The increase in the number of feedback problems will increase.The current technical level cannot completely use a computer to handle all feedback problems.A courier customer service and a computer must jointly handle and resolve these problems.The courier customer service improves customer satisfaction with the courier industry.It plays an important role in retaining customers,reducing circulation costs,and is an indispensable part of every courier company.The performance of express customer service is different from the performance of general functional departments,and its evaluation has become a major problem for express companies.Therefore,it is particularly important to establish a scientific,reasonable and fair evaluation system for express customer service.This paper takes the performance evaluation of express customer service as the research content,and applies data warehouse technology and data mining technology to it.During the establishment of the data warehouse,the process of massive data from the original data source to the data warehouse was analyzed,the data pre-processing process,methods and processing rules in the performance evaluation system were introduced in detail,and the data warehouse structure was designed;in the data mining algorithm Here,the decision tree algorithm has good classification accuracy and fast classification ability.According to the data characteristics of the express customer service,the decision tree algorithm is used to mine the historical data of the express customer service.The existing decision tree algorithms have the following characteristics in terms of computational complexity and classification attribute selection: Insufficient,this paper proposes a decision tree algorithm based on coordinated measurement information entropy to construct a decision tree model for express customer service data.Simulation results verify the superiority of the algorithm.The final model display shows the complexity and differentiation of the split attributes of the model.The problem that the performance evaluation data of express customer service changes constantly with time.By combining the Bayesian algorithm and the decision tree algorithm to solve the problem,an incremental decision tree algorithm based on the Bayesian node is proposed.The comparative analysis of the experimental results shows that Bayesian nodeincremental decision tree algorithm for The rationality of the historical data processing modeling of delivery service;based on the data processing and algorithm design optimization,this paper designs a complete express delivery service performance evaluation system,which has the main functions of data management,algorithm import,and data modeling,and has passed the basic Functional and performance tests verify that the system meets design requirements.The purpose of this paper is to transmit valuable information to management personnel in a short time,to assist management personnel in the express delivery industry to assist decision-making,and to enable management personnel to formulate targeted policies based on the analysis results to manage and coordinate customer service.The results of this paper are conducive to proposing effective management methods suitable for customer service work and suggestions for improving customer service performance,etc.,and have good theoretical value and practical significance.