Holder, Oliver (2020) Training Machine Learning Models to Automatically Identify Technical Debt in Issue Trackers. Bachelor's Thesis, Computing Science.
|
Text
bCS_2020_HolderO.pdf Download (560kB) | Preview |
|
![[img]](https://fse.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
Toestemming.pdf Restricted to Registered users only Download (92kB) |
Abstract
Technical debt is a wide spread phenomenon in the software world. It occurs when developers make wrong or unhelpful choices, often in the form of short term solutions, which increases the amount of work in the long run. This thesis aims to study machine learning techniques for the prediction of technical debt in issue trackers. Moreover, the use of convolution neural networks allows for the extraction of n-gram phrases that identify the presence of technical debt. Issue tracker repositories were used as data sets. In total 5335 tickets were annotated, used for model training, and had key words extracted. The evaluation of the models shows that convolution neural networks slightly outperform other traditional models in this application.
| Item Type: | Thesis (Bachelor's Thesis) |
|---|---|
| Supervisor name: | Avgeriou, P. and Soliman, M.A.M. |
| Degree programme: | Computing Science |
| Thesis type: | Bachelor's Thesis |
| Language: | English |
| Date Deposited: | 20 Aug 2020 16:25 |
| Last Modified: | 20 Aug 2020 16:25 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/23152 |
Actions (login required)
 |
View Item |