Prediction method for incoming call traffic of power client service center

Abstract

The invention relates to a prediction method for the incoming call traffic of a power client service center. The method comprises the following steps that 1) telephone service content information is analyzed, types of telephone services are arranged, and main problems accepted by the client service center are concluded; 2) on the basis of historical data of 3 to 5 years, noise points in the data are determined, reasons for sudden increase and decrease of the telephone traffic are analyzed, and key influential factors and influence values of the factors are confirmed; and 3) the change trend of the incoming call traffic is analyzed according to a time period, periodic rules of the telephone traffic are extracted, and a prediction model of the telephone traffic is established. The prediction method can be used to predict the telephone traffic effectively, the prediction accuracy is improved via a historical big data sample and multi-level influential factor analysis, the utilization rate of present telephone service representatives is maximized, the perfecting management level of a power company is improved, the cost of telephone construction is reduced, and response time of service demands of clients is shortened.

Description

technical field [0001] The invention relates to the technical field of communication, in particular to a method for predicting incoming call traffic of an electric power customer service center. Background technique [0002] The 95598 customer service center of the State Grid Corporation has achieved national customer service concentration. Compared with the previous distribution mode in various provinces, the number of agent representatives and the scale of service customers are unprecedentedly large. In this mode, both in terms of customer service quality assurance and reasonable agent scheduling, the difficulty increases exponentially. How to effectively control the cost of agent resources on the premise of ensuring the shortest waiting time for customer calls is the primary problem facing 95598 customer service. How to schedule agent resources first depends on how many incoming calls will be made by customers who will use electricity in the future. At present, the cust...

AI Technical Summary

Problems solved by technology

[0005] (1) The summary analysis of the influencing factors of dialogue traffic relies on manual experience, and the real influencing factors have not been analyzed through big data to verify manual experience;

[0006] (2) Because it is impossible to scientifically measure the impact of various influencing factors on traffic forecasting, the traffic forecasting model excludes other influencing factors except temperature, weather, and air temperature factors, and the prediction accuracy is low;

[0007] (3) The influencing factors of traffic forecasting rely on manual summary analysis, and the forecasting model cannot be automatically corrected after the summary analysis, and the actual forecasting relies on the manual intervention of forecasters

Images