Bug Prediction In Large Scale Software Development: Technical Report

Soldan, Robert (2022) Bug Prediction In Large Scale Software Development: Technical Report. Undergraduate thesis, Dublin, National College of Ireland.

Preview

PDF (Bachelor of Science) Download (1MB) | Preview

Abstract

Software has grown more complex, and the cost of fixing software bugs is increasing. The ability to predict the occurrence of bugs after the contribution of a developer, or
to predict what types of bugs that contribution might cause can be an asset to any company which develops software at a large scale.

This project explored the possibility of predicting a bug occurring after a contribution by using the data from the title, body, and tags of the pull request (PR). Additionally, the same data was be used to predict the category of a bug.

The data set was scraped from VSCode’s GitHub repository, and it comprises of pull request information and issues information. Issues were assigned to pull requests according to their dates, issues being after pull requests. The analysis used three machine learning models to perform the predictions: Random Forest, Decision Trees, and Linear Support Vector Machine (lSVM).

The results did not show any significant levels of accuracy in the predictions for the models tested. This may have been due to inconsistent labeling across the project, and under sampling when testing for the bug categories.

Item Type:	Thesis (Undergraduate)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software T Technology > T Technology (General) > Information Technology > Computer software
Divisions:	School of Computing > Bachelor of Science (Honours) in Computing
Depositing User:	Clara Chan
Date Deposited:	14 Sep 2022 11:54
Last Modified:	14 Sep 2022 11:54
URI:	https://norma.ncirl.ie/id/eprint/5766

Actions (login required)

View Item