Research On Visual Assessment Of Service Quality Of Operators' Installation And Maintenance

With the development of communication technology,optical cable network has been widely used,and the competition among operators has gradually shifted to the field of optical cable network.Therefore,the last link in the construction of optical network access,the customer-oriented installation and maintenance services,is increasingly valued in the internal management system of operators.The installation and maintenance process generates multidimensional operational data in the operator's operational management system,such as the time of the order was initiated,the installation duration,the address,and the success rate of installation.By analyzing the data of the installation and maintenance process,the analyst can get a capability condition of the installation and maintenance personnel,help the operator optimize the management strategy,and improve the quality of the installation and maintenance service.At present,operators collect customer evaluation score data and manage maintenance personnel through statistical methods.However,this method of management lags behind,and the next step of adjustment can not be made until the installation and maintenance service has produced bad perception.Other dimensions of data have not been effectively used,and the classification of the installation and maintenance operation data are also lack of rationality.The results are greatly affected by the analyst's subjective awareness,and it is difficult for the analyst to make a reasonable explanation.In view of the above problems,this thesis constructs a service quality evaluation model,uses the data of the installation and maintenance process to estimate the service quality of the maintenance personnel,and achieves the support for the operators to rationally formulate management strategies.It mainly focuses on the following three aspects:1.In view of the lag problem in current service quality evaluation methods of operators,this thesis,based on BP(Back Propagation)artificial neural network theory,designs a method to estimate the service quality of installation and maintenance.The evaluation model is constructed by selecting and collating multi-dimensional data such as the non-installation time on the day,the total number of points scheduled on the day,the timeliness of the first response,the installation time and the number of repairs.Comparing the change trend of error values of four different training algorithms using visual analysis method,assisting to select the optimal model,and realizing the service quality prediction and evaluation for maintenance personnel.2.To solve the problem that operators' current statistical-based data classification methods are greatly affected by analysts' subjective awareness and are difficult to make reasonable explanations,this thesis designs a classification method of installation and maintenance satisfaction evaluation data and related data by using cluster analysis algorithm,and explores the correlation between these data in the process of cluster analysis by visual analysis method,and optimizes the classification and formulation of management strategies for operators.3.In order to solve the problem of scattered and lack of integration of data analysis of operators,this thesis designs a visual analysis system for service quality evaluation,which implements visual analysis of service quality evaluation,clustering analysis and algorithm comparison through multiple options and multiple view interfaces.This thesis validates the validity of the above system using 1022 data of the installation and maintenance process in C City,and selects 20 respondents to participate in the system evaluation,and finally verifies that the system can effectively support operators to enhance the value mining of operational data,and better carry out the operation and maintenance management work.