Performance Evaluation of Call Center Agents by Neural Networks

The call center provides customer services to the customer of a company. Call center agents play an important role in such services. To ensure the quality of customer service, agent training and evaluation are essential. Usually, agents are monthly evaluated by their supervisor. Nevertheless, an objective evaluation standard is desired. Twenty three quantitative indicators for call center operations are presented and discussed in the literature. On the other hand, a call center operation company also provides its internal evaluation criteria for this research. We then decide a set of performance indicators using feature selection methods. The selected performance indicators are used as the input to the backpropagation neural network and the performance evaluation score in 100 point scale is the output. A set of one-year real data is used for training and testing the backpropagation neural network. The average error and the correlation coefficient between the predicted output values and the real evaluation scores are 2.78 and 0.9821, respectively, which are quite good. After erroneous data are fixed, the prediction results are further improved to 1.97 and 0.9907. Finally, the appraisal grades based on the predicted performance evaluation scores are compared with the original grades.