ABSTRACT
The method of classifying requirement documentations into different levels of quality is mainly based on COSMIC functional size measurement[2]and relies on recognizing function[3] and data transfer. In recognizing data transfer[4], it needs to recognize function, sub-function and data type, which are closely related to the documentation quality. Through this method, each functional process can include not only the size of functional point, but also the level of documentation quality. Analyzing the measurement results will help to improve the documentation quality and describe the reliability of measurement results. If the level of the documentation is quite low, it means that the deviations of the measurement results from
this functional process is too large, therefore, the reliability of this functional process is low, and vice versa. In the mapping stage, classifying requirement documentations into different levels of quality and providing feedback to the organizer of requirement documentations to constraint quality control play an important role through the whole measurement process. In the measurement process, firstly, we classify requirement documentations into different levels of quality of functional process. After the measurement of all the functions process is completed, we calculate the weight coefficient of each functional process in the whole process, convert the level of documentation quality into value, and apply formulas to calculate the mean of the result, and then convert the result value into the level of documentation quality, which is the level of quality of the required documentation.
