A Survey on Software Bug Prediction Practices: Tools and Techniques
VFAST Transactions on Software Engineering · 24.06.2026 · 14 (2) · 298-317
DOI: 10.21015/vtse.v14i2.2440 ↗Kurz gefasst
Fachartikel in der Zeitschrift VFAST Transactions on Software Engineering, online erschienen am 24.06.2026. Verfasst von Amna Ijaz und 4 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben.
Abstract (Originalsprache)
If using bug prediction tools can help increase software quality and reduce maintenance costs, there is a huge difference between the tools that are developed and the tools that are used in the real world. Despite the advances of machine learning and automated analysis, most development teams still do some things by hand because these tools are prone to false alarms, are difficult to integrate, and don't seem to fit within existing workflows. We set out to understand the challenges and the organizational delays that are holding teams back from applying the tools in practice and why bug prediction research is not permeating to the real world. The objective was to provide some practical guidance for researchers on what developers actually need. To get a glimpse into how P@SHA-certified software developers in Pakistan utilize bug prediction tools, challenges faced by them and what could be improved, we surveyed 95 software developers from P@SHA-certified companies in Pakistan. We considered the adoption rates, the effectiveness of the tools in use and the integration into the current processes. This study introduces a conceptual adoption framework as a result of the practitioner's feedback, focusing on the factors that are significant for the software teams in the decision making process if they adopt bug prediction tools. We discovered that 47.4% of teams continue to use manual code reviews, while only 28.4% are working with machine learning tools. While popular tools like Checkmarx (28.4%) and ESLint (27.4%) showed moderate effectiveness, but 46.3% of people had neutral opinions regarding tool performance. The most common suggestion enhancement was improved integration with IDEs and CI/CD systems—34.7% indicating this was their highest priority. In order to bridge the gap between academia and real-world software development, we need bug prediction tools within the development environment and are understandable by developers.
Zitieren
APA
Ijaz, A., Fatima, M., Qamar, N., Nazir, F., & Sabahat, N. (2026). A Survey on Software Bug Prediction Practices: Tools and Techniques. VFAST Transactions on Software Engineering, 14(2), 298-317. https://doi.org/10.21015/vtse.v14i2.2440
Vancouver
Ijaz A, Fatima M, Qamar N, Nazir F, Sabahat N. A Survey on Software Bug Prediction Practices: Tools and Techniques. VFAST Transactions on Software Engineering. 2026;14(2):298-317. doi:10.21015/vtse.v14i2.2440
BibTeX
@article{Ijaz2026,
author = {Ijaz, Amna and Fatima, Maha and Qamar, Nosheen and Nazir, Faria and Sabahat, Nosheen},
title = {A Survey on Software Bug Prediction Practices: Tools and Techniques},
journal = {VFAST Transactions on Software Engineering},
year = {2026},
volume = {14},
number = {2},
pages = {298--317},
doi = {10.21015/vtse.v14i2.2440}
}
EUVIMED