Loading...

A Code-Review Facilitator System According to Contextual Characteristics

Shateri, Pedram | 2023

12 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56873 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Habibi, Jafar
  7. Abstract:
  8. Manual code review, essential for software quality, suffers from time constraints and repetitive tasks. This thesis investigates using large language models (LLMs) with prompt engineering and in-context learning to automate aspects of the process. By leveraging an LLM's generalization capabilities, we aim to achieve automation with limited resources and minimal pre-training. Due to the development of large language models and in-context learning capabilities, our proposed approach is to add contextual information relevant to code review to the model input. Our approach focuses on providing context-specific samples and documents related to the reviewed code, enabling the LLM to learn from the context and generate results which is after review code. This contextual-information is actually taken from GitHub projects and Stackoverflow. GitHub information is for using high-quality and up-to-date code, and Stackoverflow's data is mostly used for natural language information and guidance. For the model part, large language models have been used .Evaluation using CodeBleu criteria showed a 15% improvement in performance compared to baseline models, demonstrating the potential of our method to boost code review efficiency and effectiveness. This research opens avenues for exploring the broader applicability of LLMs in automating software development tasks
  9. Keywords:
  10. Code Review ; In-Context Learning ; Retrieval-Augmented Generation ; Prompt Engineering ; Large Language Model ; Software Quality Assurance

 Digital Object List

 Bookmark

No TOC