ROSE: Transformer-Based Refactoring Recommendation for Architectural Smells

TL;DR

This paper presents ROSE, a transformer-based system that recommends refactorings for architectural smells in software, achieving high accuracy and outperforming baselines, thus advancing automated repair suggestions.

Contribution

It introduces a novel application of transformer models for refactoring recommendation, fine-tuned on a large dataset of real-world Java projects, demonstrating effective repair suggestions.

Findings

CodeT5 achieves 96.9% accuracy and 95.2% F1 score.

Transformer models outperform traditional baselines.

The approach effectively bridges smell detection and repair.

Abstract

Architectural smells such as God Class, Cyclic Dependency, and Hub-like Dependency degrade software quality and maintainability. Existing tools detect such smells but rarely suggest how to fix them. This paper explores the use of pre-trained transformer models--CodeBERT and CodeT5--for recommending suitable refactorings based on detected smells. We frame the task as a three-class classification problem and fine-tune both models on over 2 million refactoring instances mined from 11,149 open-source Java projects. CodeT5 achieves 96.9% accuracy and 95.2% F1, outperforming CodeBERT and traditional baselines. Our results show that transformer-based models can effectively bridge the gap between smell detection and actionable repair, laying the foundation for future refactoring recommendation systems. We release all code, models, and data under an open license to support reproducibility and further research.