Dekker, Arjan and Maarleveld, Jesse (2022) Mining for Architectural Design Decisions in Issue Tracking Systems using Deep Learning Approaches. Master's Internship Report, Computing Science.
|
Text
mCS_2022_DekkerAJ&MaarleveldJ.pdf Download (15MB) | Preview |
|
![[img]](https://fse.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
toestemming.pdf Restricted to Registered users only Download (123kB) |
Abstract
Issues in issue tracking systems can contain useful information about the architecture of a project, in the form or so-called architectural design decisions. A major challenge is identifying issues containing such architectural design decisions. One possible way to identify such issues is using deep learning. We used multiple deep learning approaches to detect and classify issues that are rich in architectural information. We experimented with a number of different types of neural networks, as well as multiple ways of generating feature vectors. We compared our methods with the current state of the art, which currently consists of classical machine learning techniques. We found that deep learning techniques can achieve performance similar than the current state of the art machine learning techniques. The main benefit of deep learning techniques is that they seem to be more generalizable in two different ways. Their performance generalizes well to different project, and to other datasets. This shows that deep learning is a promising technique for automatically mining architectural knowledge from issue trackers.
| Item Type: | Thesis (Master's Internship Report) |
|---|---|
| Supervisor name: | Soliman, M.A.M. and Avgeriou, P. |
| Degree programme: | Computing Science |
| Thesis type: | Master's Internship Report |
| Language: | English |
| Date Deposited: | 09 Sep 2022 11:52 |
| Last Modified: | 09 Sep 2022 11:52 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/28689 |
Actions (login required)
 |
View Item |