An Issue-driven Requirement Management Approach in a Wiki-based Collaborative Environment

• Main Authors: Sheng-Wei Tsai, 蔡升偉
• Other Authors: none
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/99882237258985624037

碩士 === 逢甲大學 === 資訊電機工程碩士在職專班 === 97 === Requirements engineering plays an important role during software development. Even though a software system may deliver powerful functions, it is still not a successful product if user expectations can not be satisfied. To develop a software system that meets user requirements, one of the most important things is to facilitate the communication between the stakeholders in the project. On the other hand, nowadays the development team is usually distributed in different locations; therefore the communication is done through the internet tools such as email, MSN, or blog. However, these tools can’t help stakeholders manage requirements, track issues and test their system. Some companies may build their own specific environment or product for communication and requirement management, but cost is too high. To provide a distributed and flexible platform for requirement engineering in low cost, we propose an IRTMS methodology and wiki-based environment to address the problems. Issue, Requirement, Test, Module, and Schedule are five major concepts in our method to lead the system development. The communication between stakeholders starts from raising issues, discussing issues and ends with archiving issues to meet requirements. Since all issue discussions are recorded in the wiki, design rationales are well-understood and the conflicts are reduced. While handling requirements, their related test cases are established on the basis of test-first strategy. Developers and users can easily collaborate to refine and define the test case in the wiki environment. Once the test cases are defined, all following modules will be tested in an automatic manner. Moreover, the schedule for each requirement is defined in the wiki environment. Therefore, it can be easily tracked, monitored and revised. In the thesis, we evaluate our approach by simulating a real project execution between Feng Chia University (FCU) in Taichung and A-University (alias name for privacy) in Taipei, where FCU is the development team and AU is the customer. Through the experiment, we can find our IRTMS method is contributive to distributed project management.