Research On The Optimization Of Large Value-adde D Service System Based On The Defect Data Analy Sis

With the increasing of the market competition, the changing of the business requirements, Lead to software version need to constantly upgrade tomeet user expectations. And also can achieve a smooth version upgrade,do not affect the user experience. It’s a very big challenge for continuoustesting working to this kind of all-weather running large, complex value-added service system. In this paper, the author selected the practical work ofthe cardholder value-added service platform project. First of all, finding and combing the problems of defect classification and analysis stage. To solve these problems, corresponding measures should be taken. Finally, evaluate the implementation of the expected effect.In this paper, based on the production defects data in the database, First, according to the different subsystems, stage, rank, type and other general properties, calculate out the distribution of the defects, the Important defects can be found according to the statistics results and preliminary countermeasures are proposed. In addition, orthogonal defect classification methodof IBM was introduced. On the basis of defects in the general propertiesand adding the ODC characteristic properties, such as Activity、Trigger、Impact、 Target、Defect Type、Qualifier、Age、Source. According to the ODC attributes classification standard, need to reorganize the defects data and prepare for the subsequent analysis of ODC. In defect analysis part, first introduced the root cause analysis methodto analyze the cardholder service platform project, Drawing the fishbone diagram, and Summarizes the following7problems: code, platform, operation, monitoring filtering rules, repeated or invalid issue, problem can’t be reproduced. and then respectively analyze these problems to find out the rootcause, and put forward improving plan, evaluate the results of feedback. Because of the relatively high cost of root cause analysis, A more practicalorthogonal defect analysis method is introduced. the ODC method integrating classification and analysis, On the basis of defect classification before.according to the need analysis of the different goals, selecting different ODC attributes, using software to output a variety of different dimensions ofanalysis diagram. Such as in order to evaluate the design of the productstability, and find out the shortage of the design and code, select the Defect Type and Qualifier properties to analysis, find the unstable factors, andtake action to solve the problem.Due to the released version is continued, The classification and analysis of defect data is also need to constantly, Optimize work also needs to keep going. Thus introduces the ideas of the Deming’s PDCA continuous improvement, and try to through it to the cardholder service platform projectclassification and analysis work, and formation of the work flow, standardization, and to establish a long-term and reasonable defect management mechanism, and gen ralize to the similar value-added service system.