Ever since the early 90¡¦s the global software market has been growing at a rapid speed. Along with the development of Internet and electronic commerce, the demands for software have been increased dramatically and Taiwanese software firms are encountered serious global competition too. To stay competitive, the Taiwan software industry has to improve the quality of its software products in order to compete in the international IT arena. The present best option is for the Taiwanese software firms to achieve the Software Engineering Institute¡¦s Capability Maturity Model certification. This study compares the major differences on the requirement management process between CMMI level II and level III. The differences include key process¡¦s purpose, activity, and artifact. Two real-word cases were used to illustrate the example and concept. These results provide a practical bases and better understanding for the Taiwanese software firms when they implement CMMI level II or III.

Capability Maturity Model Integrated (CMMI) developed by Software Engineering Institute of Carnegie Mellon University is aimed to improve the productivity of software development. From CMM to CMMI, they have been adopted worldwide. But the large numbers of unsuccessful and/or painful implementations have proved these implementations easier said than done. The software industry in Taiwan faces the challenge of global competition. Continuous improvement of productivity becomes inevitable. To achieve this goal, recently, government and non-government organizations together promote the CMMI. This thesis focuses on the Decision Analysis and Resolution of CMMI Level 3. Through the design of related methods and the development of software tools, the decision process can be more structured and more effective. Potential applications include evaluation and selection of software and hardware, outsourcing partner, tools, risk resolution alternatives or people. The typical Go/No-Go decision can also fit into it. The decision analysis methods included in this research are Multi-attribute Decision Making and Analytic Hierarchy Process(AHP). In addition, alternative screening and group decision making can be integrated into the above methods.