Field-oriented Software Components Evolution Metrics Research

With the increasingly development of computer application and software engineering, software system would become old-ageing gradually daily. Just as other systems, we named the system which is out of date as legacy system. Software evolution has become an important characteristic in the software life cycle. Today, software development technology continues to improve, software development scale and complexity is also increasing. In order to reduce development risk, lots of software companies have improved the software process through the process of measurement. Therefore, the measurement has become a basic practice of the software engineering. On the other hand, Component based software engineering is one of the most effective solutions to improve software development quality and productivity. However, in addition to component composition, component evolution is another major problem. Therefore, component evolution has become a new field in Component-based software engineering.First of all, in this thesis, related concepts about software evolution process, software metric and software component are analyzed. We have pointed out the difficulties in the actual measurement process of Software Company, and put forward some suggestion to cope with it. Next, using the thought of GQM for reference, we have put forward a component evolution metric model facing to field (F-CEMM), in order to support component evolution metrics of the software system. In the implementing measures process of software companies, different interest of roles have different understanding about the evolution process of measurement, and we have analyzed the relationship in detail depending on the different information needs. Secondly, the thesis analyzes the different metric goal, and we have deeply studied selection of the metric elements. Then list some base information measurement and derived measurement, which can help software system to value the progress of the component evolution and support it in decision-making. Next, we have dentally analyzed about data acquisition, and analyzing of metric results, and so on. Finally, the thesis gave an application of this model, in order to effectively control the process of component evolution.In summary, the research of this thesis effectively improved the quality of component evolution in software systems. By measuring and evaluating, it hopes to effectively control the component evolution, and finally it could improve the productivity of software companies in practice.