University of Twente Student Theses
Using LLMs to aid developers with code comprehension in codebases
Reefman, K.H.A. (2024) Using LLMs to aid developers with code comprehension in codebases.
|
PDF
3MB |
| Abstract: | Developers spend significant amounts of time on comprehending code, with research suggesting developers spend up to 70\% of their time on code comprehension. This is worsened further if there is a lack of proper documentation. With the help of Large Language Models, this could be sped up greatly. However, current research on this topic is limited in scope, mostly focusing on an educational context. In this work, the effectiveness of LLM tools is evaluated in a real-world setting, with the usage of locally running models. Results show that such tools can help developers quickly narrow down where to look in the code when solving tasks. Additionally, results show that developers can get a good low-level understanding of the code without the aid of an expert on the codebase. The nature of this research, with its usage of locally running models, allows for several directions for improvements in future work. |
| Item Type: | Essay (Master) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science MSc (60300) |
| Link to this item: | https://purl.utwente.nl/essays/103120 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page