Method Of Component Impotance Measurement In Software Evolution Environment

Software system has gradually evolved into the combined delivery of services and components,and has been constantly adjusted and expanded for the sake of social development.The scale of software system has increased so much,and the structure has appeared in different levels,different granularity and multiple integration ways.It is common in software systems that a series of complex changes in software systems that are gradually changing to an ideal form are software evolution.However,with the development of the function and scale of the software system,the mastery and control of the software evolution have become more complicated and more difficult.Although traditional measurement methods show some characteristics of software evolution,these methods all fall into the complex details of software structure in earlier time generally,which is not enough attention to the macro aspect of software evolution,and it is difficult to grasp the software structure overall and comprehensively.In view of this problem,this paper holds that measure important components can not only save the cost of evolution when complicated and large-scale systems evolved,but also help to better grasp the evolution process of these important components in the evolution of the software architecture,reduce the risk of evolution and avoid wasting the evolutionary resources at the same time,on the other hand,focus on monitoring and managing those activities and components that are more difficult to control in software evolution.Therefore,this paper worked out a method to grasp the software architecture in the macro aspect,and measure the importance of the components in the complex software system,and provide the basis and support for the evolution of the complex software system.Worked out a software architecture directed graph model is proposed.This model based on software architecture,takes components of software system as nodes of directed graph model,and the dependency relationship among components as directed edges.On the basis of the directed graph model of the software architecture,the betweenness and the total dependence are calculated,and as the data base and support provided for the comprehensive measurement of the important components.Combining the total dependence of the betweenness and the component,measured the importance of the component synthetically,analyzed the correlation of these two indexes,and the correctness of the measurement method is verified.In summary,the work of this paper provides a powerful basis for measuring the important components of software system in software evolution.It will save time and resource cost of software evolution,improve the quality and efficiency of software evolution,and supplement the shortcomings of traditional measurement methods in mastering the macro characteristics of software architecture.