An Application Of Neural Network To Service Quality Management Of Call Center

CSRs(Customer Service Representatives)play a critical role in providing an interface between customer and company. They are the frontline staff in customer interaction. Therefore,the performance of CSRs will affect the overall service quality of the call center. The paper applies Neural Network techniques to the problem for predicting the service quality of CSRs in call centers and analyzes the characteristics of the input sensitivity.Firstly, the paper introduces the characteristics of service quality management of call center, the operation flow and the present status of service quality management of telecommunication call centers. Then, the paper applies the Multiple Linear Regression Model to predicting the service quality of CSRs and due to the complexity and nonlinear relation between the inputs and outputs, the model is unfitted. The traditional method tools have not met the need. Consequently, taking advantages of artificial neural network, the paper sets up models with the well-known Back Propagation neural network and probabilistic neural networks approach. The empirical study is based on the performance data actually collected in a call center of a large telecommunication company. The prediction accuracy of the model are 72.86%(BP)and 84.32%(PNN). The result of input sensitivity indicates that product is very important to predicting the performance of CSRs, which reveals that CSRs in some products have more opportunity to far exceed than the CSRs in other products. First call resolution is important too. Finally, the paper puts forward the suggestions to improve the service quality management of CSRs. We hope that prediction will help to lessen the calls that evaluators reviewed in order to save cost and inspirit CSRs.