The State Grid Service Customer Satisfaction Research

In electric power system reform especially sell electricity side open,against the background of the power purchase and sale of unified the traditional power mode is gradually being broken,and customer service quality is vital for power grid companies to become.Improve the level of customer service management and customer satisfaction,is to realize the precise marketing,precision and guarantee the premise of risk control,differentiated services,is also the important tool to improve the corporate image and the economic efficiency of enterprises.95598 is the main window of power system geared to the needs of customers,the multiple channels such as telephone,SMS,website at the same time of providing service for the customer to generate a large amount of data,including structured and unstructured data,these data to assess the customer service quality,customer satisfaction provides a data basis.Traditional satisfaction analysis was based on numeric data,the index system of building satisfaction,satisfaction with the method of modeling and analysis of satisfaction evaluation,and predecessors' studies,in the article,according to the work of its electric power single data for text analysis,this paper puts forward a set of customer service satisfaction analysis framework based on text mining,through the framework,can not achieve satisfactory work order recognition and not satisfied with the reason of recognition.Using the analysis framework,power grid companies can realize semi-automatic work order for customer service,text analysis,greatly reduce the labor costs,improve the service efficiency.This article is to research the power customer satisfaction through our PLsql platform for its customer service work order data is extracted,and choose the suitable degree of satisfaction analysis of work in a single data,using the corresponding rules for data preprocessing,mining method adopted in Chinese word segmentation model,emotion recognition and themes,for not satisfied with the repair order and effective to identify the cause of dissatisfaction,showed its power service,provide targeted SGC good prevention and correction advice,the thesis of the whole process is a PLsql and R software.In this paper,the effect of text mining on the recognition of unsatisfactory work orders and unsatisfactory reasons is significant,and the accuracy of job recognition is 90%,and the coverage is less than 90%and the accuracy is 93%.At the end of this paper,we put forward some reasons to improve the satisfaction of customer service,and strive to strengthen the construction of large data platform.