Enhancing software quality through predictive analytics: A deep learning approach to defect prediction and prevention
Department of Software Quality Engineering, Apple Inc, USA.
Research Article
Global Journal of Engineering and Technology Advances, 2024, 19(01), 212-221.
Article DOI: 10.30574/gjeta.2024.19.1.0065
Publication history:
Received on 01 March 2024; revised on 10 April 2024; accepted on 13 April 2024
Abstract:
In the case of modern applications, software quality is essential because it reflects how reliable, secure, and maintainable the application will be. Manual review and rule-based testing are traditional, but not robust, methods of detecting and preventing defects in software systems, and they tend to fail when the complexity of the software systems grows. This paper investigates how software quality can be improved by early defect prediction and prevention using predictive analytics and deep learning. The base of machine learning powered predictive analytics uses historical defect data to detect patterns that could signal a potential weakness in software. One of the attractive aspects of deep learning is that, due to automatically learning complex representations, it can achieve even higher accuracy in locating defects during software development and automating software quality assurance processes. This paper discusses how different architectures of deep learning such as convolutional neural networks (CNNs), recurrent neural networks (RNNs) and transformers are applied to defect prediction. In addition, it provides some initial insights for implementing deep learning in the domain of software quality assurance, including the lack of data availability, high computational costs and model interpretability. The paper concludes with future research directions to improve defect prediction accuracy, to increase model transparency, and to answer ethical questions such as whether it is ethical to elevate model predictions over human decisions in the practice of software engineering.
Keywords:
Software quality; Defect prediction; Predictive analytics; Deep learning; Machine learning; Software engineering; Software defects; Quality assurance; Automated testing; Neural networks
Full text article in PDF:
Copyright information:
Copyright © 2024 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0