Call Center Technology Research Based On J2EE For Scheduling System

In the fierce market in which competition keeps increasing day by day, it's most important for a call center to cut the operation cost, reduce lost calls and improve the service quality. The subject of this paper is on how to realize the key aims, that is, how to predict calls correctly, how to schedule staff's work effectualy by their skill,how to better service quality and how to raise the efficiency of service and the rate of customer satisfaction on the basis of a fixed cost budget.Firstly, this paper analyses the disadvantages of scheduling mannually which are as follows:it's complicated to compute, unavoidable to consume time, hard to get data, hard to find the call-changing tendency, hard to schedule staff's work by their skill, hard to satisfy staffs and hard to analyse staffs choices, etc.Secondly, J2EE distributed architecture is discussed. An architecture has been put forward to integrate the system with information based on the data source layer, cored on the business layer and accessed by the view layer.Next, how to schedule in a predicting way and how to forecast calls and human resource requirements are also discussed. Based on statistics of history data of calls, a model which could predict calls in a short or medium or long term is put forward. With Erlang C algorithm, this model predicts queue time, staffs workload and best staff-schedule, computes the number of staffs to meet the demands of service level as well as getting a method of revising the forecasts. In the meantime, the division of function modules for scheduling system is introduced and business forecasts, requirements for workforce forecasts, scheduling results, asking for leave, modifying schedules and result-monitoring have been designed in detail.This research is carried out on the basis of the system design and development of the scheduling system for the call center of a certain mobile operator. The result proves this scheduling system is scientific and practical.Finally, this paper also points out the problems about the forecast algorithm and business functions which need to be further discussed and solved.