Research And Implementation Of Quantitative Management Tool In Agile Software Development

As the demand for software products grows because of the rapid development of information technology, software products have stretched into every corner of society, control software on small electronic products, system software on the Space Shuttle, as well as a variety of trading systems, e-commerce systems, management systems, and so on. Huge market demand makes software industry has been developing rapidly, traditional development model and process in the new commercial and technological environment has begun exposes the weaknesses of adaptability, more and more development teams and organizations apply agile development methods, agile software development uses timeboxed iterative and evolutionary development, adapative project plan, promotes evolutionary delivery, has high responsibility and adaptability to changes.Agile project management is a new project mangement pattern which aims at the agile software development. Compare with the traditional project management pattern, APM has more flexibility and maneuverability. However the flexibility also means confusion to some extent. As a matter of fact, in real APM, the requirements may vary frequently, which in turn increase the risk of whole project, besides, the project plan, development progress, and testing program will all be affected. Therefore, the existing requirement management process needs to be improved and optimized in order to minimize the negative effects brought by the requirement changes.This paper focues on the requirement management process in actual APM, extract, analyze, and improve the process, combine the requirement management with test case management and bug management together to build a centralized integrated management system. With this effort, the workload of managing and tracing requirements, test cases and bugs can be reduced dramatically, and make the work be more effective and efficient.Meanwhile, the existing software quality measurement theory and method are mostly software products oriented, which neglect the metrics of product developers themselves. People usually use code level software quality metrics to mesure the quailty of developer's work indirectly. This method is widely used. However, we should show more respect to ones historical performance, it will be more practical if the human performance measurement bases on the quantitative data, in addition to qualitive analysis. This paper presents a statistics-based method for staff performance and capability measurement which aims at helping project managers to do a moreprecise evaluation when they need to measure the performance of developers in the team. These indicators and statistics are also expected to balance the load within staffs, improve the overall resource utilization, and make the whole be more effective.