A search engine for finding and reusing architecturally significant code

Abstract

I am grateful to my thesis advisor Dr. Mehdi Mirakhorli. I could not have achieved this without his invaluable motivation, patience, and dedication to this work. I also thank Dr. Meiyappan Nagappan for serving on my thesis committee and his insightful comments and encouragement. Furthermore, I thank the rest of the department faculty and staff for their support through my study. I would like to express my gratitude to Joanna Cecilia for her active participation in this research work and for her knowledge and contribution for it. Also, I would like to thank my team member Raghuram Gopalakrishnan for assisting in conducting experiments for my research. Finally, I would like to thank the Iraqi High Committee of Education Development for the full scholarship to complete my Master study.

Improving code summarization by combining deep learning and empirical knowledge (S)

“…We use 85% of the dataset for training, 10% for validation and 5% for testing. We has published all of our data and codes in Trustie 6 .…”

“…On one hand, software developers use code summaries to understand the specification of API. On the other hand, code retrieve system relies on code summaries for effective natural language code search [6]. Thus, it is crucial to maintain high-quality and adequate code summaries in OSS projects.…”