Improving Software Quality Using an Ontology-Based Approach

iii 谨将此论文 献给敬爱的父亲，罗孝杨 对您的怀念将伴我一生 和我的家人 母亲，蒋亚平 爷爷， 罗重秀 奶奶， 蒋金莲 外公， 蒋荣陪 外婆， 黄淑华 公公， 周垠庚 阿婆， 蒋宏英 岳母， 曾桂芳 你们的关爱是我完成此文的动力之源 iv Acknowledgements I express my deepest gratitude to Dr. Doris Carver, my supervising professor, for her guidance, patience, and encouragement throughout my graduate studies. It is an honor and a blessing for being her student. She has a great knowledge and her wisdom is second to no one. Also, her honesty and integrity is my moral model. I wish to express sincere thanks to Dr. Allyson Hoss for her advice. She was persistent in her encourage and support. She had spent a lot of time on my research. As a full time employee while trying to pursue the degree, I would express my heartfelt appreciation to my employers, Dr. the director of High Performance Computing (HPC) for their active support and fully cooperation. Also, my gratitude extends to my committee members Dr. for their advice and cooperation. I give a special acknowledgment to all the software engineers who participated in my statistical survey and gave their advices on my research. I am grateful for my parents, Mr. and Mrs. Xiaoyang Luo and Yaping Jiang as well as my sister, Yunfei who are always my source of motivation and encourage me to study by their examples. I would like to sincerely thank my mother-in-law, Guifang Zeng, who handled the tasks that were due to me. I would like to thank my wife, Yao and my daughters, Muzi and Mulan for everything. Abstract Ensuring quality in software development is a challenging process. The concepts of anti-patterns and bad code smells utilize the knowledge of reoccurring problems to improve the quality of current and future software development. Anti-patterns describe recurring bad design solutions while bad code smells describe source code that is error-free but difficult to understand and maintain. Code refactoring aims to remove bad code smells without changing a program's functionality while improving program quality. There are metrics-based tools to detect a few bad code smells from source code; however, the knowledge and understanding of these indicators of low quality software are still insufficient to resolve many of the problems they represent. Minimal research addresses the relationships between or among bad code smells, anti-patterns and refactoring. In this research, we present a new ontology, Ontology for Anti-patterns, Bad Code Smells and Refactoring (OABR), to define the concepts and their relation properties. …