Software issue reports classification using association mining

Bin Zolkeply, Mohd Syafiq 2023. Software issue reports classification using association mining. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF (Mohd Syafiq Bin Zolkeply, PhD, Thesis) - Accepted Post-Print Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (4MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (101kB)

Abstract

When a software system transitions into the maintenance phase, users often raise a significant number of issues pertaining to the system. These issues encompass a wide range of topics, including software bug reports, user experience sharing, and feature improvement requests. Therefore, it is necessary to classify them before they can be given to the appropriate developers for resolution. However, the process of manually categorising these issue reports is characterised by a significant amount of labour, a substantial investment of time, and a susceptibility to errors. Although there have been studies conducted on the automation of this process, they primarily depend on prevalent or recurring patterns found within the datasets. In instances where reports diverge from prevailing patterns, there is a higher likelihood of misclassification occurring. This thesis is driven by the motivation to present a novel approach for classifying software issue reports, drawing inspiration from the technique of classification using association mining. In contrast to the conventional approach of just mining dominant patterns from the data, the thesis revealed the efficacy of mining both dominant and weak patterns from the data. Furthermore, it demonstrated the potential of using these patterns collectively in order to categorise issue reports. The experimental results demonstrate that our novel approach, which was evaluated on benchmark datasets derived from four open source software systems, has comparable accuracy to the current state-of-the-art methods. Furthermore, our method possesses unique advantages over the existing approaches.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Computer Science & Informatics
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Date of First Compliant Deposit:	13 February 2025
Date of Acceptance:	22 January 2025
Last Modified:	13 Feb 2025 15:38
URI:	https://orca.cardiff.ac.uk/id/eprint/176160

Actions (repository staff only)

Edit Item