| Software process measurement is an effective technology to manage a software project, improve software productivity and quality. The link between items can not be considered by the existing software process measurement model a good measure of which also failed to solve the problem that quantitative prediction measurement of software process and is difficult to improve the effectiveness of measurement. So it presents a less performance in track software process and control software quality.Based on Practical Software Measurement model, this thesis propose an idea that the measurement model should consider to put the relevance of product and process into the model, this thesis gives out a design of the QVMI(Question-Variety-Metrics-Indicators) measurement information model that support custom cutting in measurement system and measurement process. The model improved the effectiveness and flexibility of measurement, by which it split the measurement information into levels, implements a statistical analysis on measurement data and provides Correlation-Coefficients associated for the product and process.To solve the problem of quantitative prediction process activities, this thesis designed the OP (Organizational-Project) measurement process model integrated by the QVMI measurement information model. The model effectively improves the software process capability by dividing the measurement into double-frame of organizational-level and project-level to achieve the software project process tracking with organizational process performance baseline, and provides a basis for quantitative prediction of software process.Based on the above studies, this thesis design and implement the Software Process Measurement Sub-System, Finally through the examples, the validity of these model in measurement is demonstrated. The results show that the theory-feasibility and technical-validity of this is correct manage software quality and improve product efficiency. |
PDF Full Text Request