| Since the beginning of the 21 st century,with the increasing convenience of information interaction,people’s demands for services have increased day by day.The quality of customer service has become crucial for power grid companies.Under the background of power big data,according to the current situation,the current 95598 network customer service center can conduct in-depth research in two key links: work order classification and efficiency evaluation of agent processing work order.First,the current work order types of the 95598 network customer service center are summarized through experience.There are too many work order business classification labels.Customer service personnel may make mistakes in the process of sending orders,resulting in problems such as untimely power transmission and serious faults,increasing the workload of staff.Therefore,it is necessary to simplify the work order classification,improve the work efficiency of staff,enhance the service quality of electric power.Secondly,KPI is used to evaluate the efficiency of service agents.If the difficulty and amounts of work orders handled by two customer service agents are different,it is difficult to judge who is better by using the original KPI statistical method.It can’t balance "quality" and "quantity".Therefore,it is necessary to improve the original KPI calculation method,and establish a scientific and fair service evaluation system,to comprehensively improve the service quality of staff.Based on the above reasons,this paper first aims at the optimization problem of the 95598 work order classification of power grid company,taking the complaint work order which can best show the customer’s demands as an example,based on the power big data,adopts the method of multi-technology fusion,and carries out the cluster integration analysis of complaint work order samples by combining the word segmentation tool and the term frequency-inverse document frequency(TF-IDF)algorithm.This paper proposes an improved scheme of the95598 work order based on the computational language method,which can improve the service efficiency and satisfaction of customers.This method can realize the segmentation of words,the construction of the feature vector model,and the similarity analysis of original complaint classification.The results show that the scheme can provide a reference for the simplification of work order classification and discuss the feasibility of work order classification and merger.Secondly,this paper establishes an Empirical hierarchical Bayesian(EHB)model to evaluate the efficiency of customer service agents in processing work orders.The model takes teamwork characteristics,the types of work orders,the number of work orders,and the difficulty of work orders into account.Finally,the model is used to get an effective efficiency scoring result.The results show that compared to the performance evaluation results of the KPI model,the EHB model is more feasible and superior in evaluating the efficiency of service agents processing work orders.This paper realizes the simplification of work order classification and fair performance evaluation of customer service staff,which has practical application value to improve the work order processing efficiency of customer service staff and improve the service quality of power 95598 at the same time. |
