International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | AI-Based Detection of Low-Quality Code Using Machine Learning Beyond Syntax-Level Analysis |
|---|---|
| ABSTRACT | Traditional code analysis tools primarily focus on detecting syntax errors and simple rule violations, but they often fail to identify deeper issues related to code quality such as poor maintainability, high complexity, and bad design practices. This paper proposes a machine learning-based approach to automatically detect low-quality code by analysing structural, semantic, and behavioural features of source code. The proposed system extracts multiple features including code complexity metrics (e.g., cyclomatic complexity), code smells, naming conventions, duplication patterns, and maintainability indices. These features are used to train machine learning models such as Random Forest (RF), Support Vector Machines (SVM), and deep learning models to classify code into high-quality and low-quality categories. Unlike traditional static analysis tools, the proposed model learns patterns from real-world code datasets and provides predictive insights into code quality. The system is evaluated using publicly available datasets from open-source repositories, and performance is measured using accuracy, precision, recall, and F1-score. Random Forest achieved the highest classification accuracy of 94.32% with an AUC of 0.9861, outperforming both SVM (91.78%) and rule-based static analysis baselines. Experimental results demonstrate that the machine learning approach significantly improves the detection of low-quality code compared to rule-based methods, contributing to automated, intelligent code quality assessment that assists developers in improving code maintainability and reducing technical debt. |
| AUTHOR | PROF. MANJULA P, SYEDA AYESHA, KHUSHNIDA HASANSAB SAYED Assistant Professor, Dept. of CS&E., Jain Institute of Technology, Davangere, Karnataka, India. UG Student, Dept. of CS&E, Jain Institute of Technology, Davangere, Karnataka, India. |
| VOLUME | 184 |
| DOI | DOI: 10.15680/IJIRCCE.2026.1405077 |
| pdf/77_AI-Based Detection of Low-Quality Code Using Machine Learning Beyond Syntax-Level Analysis.pdf | |
| KEYWORDS | |
| References | [1] Besker, T., Martini, A., & Bosch, J. (2019). Software developer productivity loss due to technical debt — A replication and extension study examining developers' development work. J. Syst. Softw., 156, 41–61. [2] Lenarduzzi, V., Besker, T., Taibi, D., et al. (2021). A systematic literature review on technical debt prioritization: Strategies, processes, factors, and tools. J. Syst. Softw., 171, 110827. [3] Chidamber, S. R., & Kemerer, C. F. (1994). A metrics suite for object-oriented design. IEEE Trans. Software Eng., 20(6), 476–493. [4] McCabe, T. J. (1976). A complexity measure. IEEE Trans. Software Eng., SE-2(4), 308–320. [5] Malhotra, R., & Jain, A. (2012). Fault prediction using statistical and machine learning methods for improving software quality. J. Inf. Proc. Syst., 8(2), 241–262. [6] Hall, T., Beecham, S., Bowes, D., et al. (2012). A systematic literature review on fault prediction performance in software engineering. IEEE Trans. Software Eng., 38(6), 1276–1304. [7] Sharma, T., & Chandra, M. (2018). Machine learning approaches for software code smell detection. Proc. Int. Conf. Inventive Research in Computing Applications (ICIRCA), Coimbatore, 978–982. [8] Palomba, F., Bavota, G., Di Penta, M., et al. (2018). Detecting bad smells in source code using change history information. Proc. 28th IEEE/ACM Int. Conf. Automated Software Engineering (ASE), Silicon Valley, 268–278. [9] White, M., Tufano, M., Vendome, C., & Poshyvanyk, D. (2016). Deep learning code fragments for code clone detection. Proc. 31st IEEE/ACM Int. Conf. Automated Software Engineering (ASE), Singapore, 87–98. [10] Guo, Z., Zhang, H., Liu, L., & Zhang, X. (2022). Graph neural networks for code quality analysis: Capturing structural dependencies beyond token sequences. IEEE Trans. Software Eng., 48(9), 3271–3285. |
