ML-Enhanced Code Refactoring Recommendations for Improving Software Maintainability

Abstract

Software maintainability is a critical quality attribute influencing the long-term sustainability, scalability, and cost-effectiveness of software systems. Traditional refactoring approaches often rely on manual inspection or rule-based static analysis, which can be time-consuming, inconsistent, and limited in capturing deeper code quality issues. Recent advances in machine learning (ML) provide new opportunities to automate and enhance refactoring recommendations by learning from large codebases, identifying complex patterns, and predicting optimal refactoring strategies. This paper investigates ML-driven approaches for generating code refactoring recommendations aimed at improving maintainability. We review existing techniques, propose an ML-based framework capable of detecting maintainability hotspots and suggesting targeted refactorings, and evaluate its effectiveness through empirical experiments on real-world repositories. Results show that ML-enhanced recommendations outperform traditional methods in accuracy, relevance, and impact on maintainability metrics. The findings highlight the potential of integrating ML into modern development practices to support developers in producing cleaner, more maintainable software systems.