| COSMIC measurement method, as one of the main-stream functional size measurement methods is widely used in business application, real-time software, as well as a mix of them. However, when carrying out the functional size measurement for the software functional user requirements,the application of the COSMIC method not only need quite a lot of measure elements identification( such as : functional user, functional processes, objects of interest, data groups, various types of data movement, etc.), but also need the separate analysis of the exceptional cases of the requirement. Therefore, the workload increases accordingly.Based on the research and use of COSMIC standard measurement method, we summed up the main workload generated section, assessed its impact on the measurement speed, analyzed the reason of the impact on rapid estimation, proposed a rapid estimation method. This method is by applying Chinese sentence constituents and speech analysis to the Functional User Requirements of software to quickly identify the types of data movement Entry and Exit, and then apply regression analysis to identify ΣEntries and ΣExits theoretical values and indicates that the software functional size Σ CFP(COSMIC function points) of the linear relationship between the value, to arrive at a method of linear regression formula. By the final predictability experiment of this rapid estimation method, it shows that the proposed software functional size rapid estimation method which is based on the linear regression analysis has some applicability.The main work of this paper is as follows:1.Studied the current mainstream functional size quick estimation methods, analyzed the basic rules, as well as the scope and practical application stage, the analysis results show that these methods are mainly used in the early stages of a software project requirements document or the initial formation.2.Presented a rapid identification data method which is based on the Chinese sentence analysis. By studying the software functional user requirement’s metrics principle of COSMIC method, we know that the identification of data movement is the Chinese Sentence analysis of the functional user requirements document. For this feature, we proposed to identify the data movement by the analysis of Chinese sentence composition and speech. This method has the following two characteristics: First, simple to apply,to start with Chinese sentence composition and speech analysis, it reduces the Professional level requirements,improves software estimation’s engineering degree, and corresponding increase the scale of software metrics working efficiency; the second is that the measurement results is more objective, the application of the method for data movement recognition require no add or delete and other operations of the requested document, excluding the subjective judgment factor’s effect of the measure person.3.Introduced the concept of linear regression analysis to the estimate of software functional size. The applicability of Linear regression analysis theory in the software estimate has already been proved, the article selected COSMIC official and 12 test cases which are already implemented in reality, it has been testified that there is a linear relationship between the regression by the analysis of test data in the way of linear regression theory, the introduction of it into the software functional size measurement is feasible. Through building a binary linear regression model of the analyzed experimental data, and then obtained final formula by the method of least squares regression.4.Predicted the software functional size rapid estimation formula’s effect in the actual estimate, which is based on linear regression Still choose the COSMIC official and the case which has been achieved in reality to apply the above mentioned rapid estimation methods, analyze and verify its effect of the software functional size estimation. Test results show that the proposed fast estimation method can significantly reduce the estimated time. |
PDF Full Text Request