Arxiv

2025

1. Arxiv

   From empirical evaluation to context-aware enhancement: Repairing regression errors with LLMs

   Anh Ho , Thanh Le-Cong , Bach Le , and Christine Rizkallah

   Under Review, 2025

   Abs DOI

   Software systems constantly evolves to adapt to ever-changing customer demands and evolving markets. During the software evolution process, code changes across versions can unexpectedly break previously working functionalities, leading to so-called regression bugs. Manually repairing regression bugs is both challenging and time-consuming, as the process often requires developers to traverse the history of the software in order to understand how the bugs were introduced and, consequently, how to repair them. Despite the challenges of manual efforts, only a limited number of studies have investigated automated program repair (APR) for regression bugs, most of which date back several years and focus on traditional APR techniques. Since then, various APR approaches, especially those leveraging the power of large language models (LLMs), have been rapidly developed to fix general software bugs. Unfortunately, the effectiveness of these advanced techniques in the context of regression bugs remains largely unexplored. This gap motivates the need for an empirical study evaluating the effectiveness of modern APR techniques in fixing real-world regression bugs. In this work, we conduct an empirical study of APR techniques on Java regression bugs. To facilitate our study, we introduce RegMiner4APR, a high-quality benchmark of Java regression bugs integrated into a framework designed to facilitate APR research. The current benchmark includes 99 regression bugs collected from 32 widely used real-world Java GitHub repositories. We begin by conducting an in-depth analysis of the benchmark, demonstrating its diversity and quality. Building on this foundation, we empirically evaluate the capabilities of APR to regression bugs by assessing both traditional APR tools and advanced LLM-based APR approaches. Our experimental results show that classical APR tools fail to repair any bugs, while LLM-based APR approaches exhibit promising potential. Motivated by these results, we investigate impact of incorporating bug-inducing change information into LLM-based APR approaches for fixing regression bugs. Our results highlight that this context-aware enhancement significantly improves the performance of LLM-based APR, yielding 1.8× more successful repairs compared to using LLM-based APR without such context.

Journal

2025

1. JSS

   EnseSmells: Deep ensemble and programming language models for automated code smells detection

   Anh Ho , Anh M. T. Bui , Phuong T. Nguyen , Amleto Di Salle , and Bach Le

   Journal of Systems and Software, 2025

   Abs DOI

   A smell in software source code denotes an indication of suboptimal design and implementation decisions, potentially hindering the code understanding and, in turn, raising the likelihood of being prone to changes and faults. Identifying these code issues at an early stage in the software development process can mitigate these problems and enhance the overall quality of the software. Current research primarily focuses on the utilization of deep learning-based models to investigate the contextual information concealed within source code instructions to detect code smells, with limited attention given to the importance of structural and design-related features. This paper proposes a novel approach to code smell detection, constructing a deep learning architecture that places importance on the fusion of structural features and statistical semantics derived from pre-trained models for programming languages. We further provide a thorough analysis of how different source code embedding models affect the detection performance with respect to different code smell types. Using four widely-used code smells from well-designed datasets, our empirical study shows that incorporating design-related features significantly improves detection accuracy, outperforming state-of-the-art methods on the MLCQ dataset with with improvements ranging from 5.98% to 28.26%, depending on the type of code smell.

Conference

2023

1. EASE

   Fusion of deep convolutional and LSTM recurrent neural networks for automated detection of code smells

   Anh Ho , Anh M. T. Bui , Phuong T. Nguyen , and Amleto Di Salle

   In Proceedings of the 27th International Conference on Evaluation and Assessment in Software Engineering , 2023

   Abs DOI

   Code smells is the term used to signal certain patterns or structures in software code that may contain a potential design or architecture problem, leading to maintainability or other software quality issues. Detecting code smells early in the software development process helps prevent these problems and improve the overall software quality. Existing research concentrates on the process of collecting and handling dataset, then exploring the potential of utilizing deep learning models to detect smells, while ignoring extensive feature engineering. Though these approaches obtained promising results, the following issues need to be tackled: (i) extracting both structural and semantic features from the software units; (ii) mitigating the effects of imbalanced data distribution on the performance.In this paper, we propose DeepSmells as a novel approach to code smells detection. To learn the complex hierarchical representations of the code fragment, we apply a deep convolutional neural network (CNN). Then, in order to improve the quality of the context encoding and preserve semantic information, long short-term memory networks (LSTM) is placed immediately after the CNN. The final classification is conducted by deep neural networks with weighted loss function to reduce the impact of skewed data distribution. We performed an empirical study using the existing code smell benchmark datasets to assess the performance of our proposed approach, and compare it with state-of-the-art baselines. The results demonstrate the effectiveness of our proposed method for all kinds of code smells with outperformed evaluation metrics in terms of F1 score and MCC.

2022

1. SOICT

   Combining deep learning and kernel PCA for software defect prediction

   Anh Ho , Hai N. Nguyen , and Anh M. T. Bui

   In Proceedings of the 11th International Symposium on Information and Communication Technology , 2022

   Abs DOI

   Software defect prediction aims to automatically determine the most likely location of defective program elements (i.e., statement, method, class, module etc.). Previous studies for software defect prediction mainly focus on exploring designing features such as source code complexity, object oriented design metrics etc. to classify program elements into two categories: (i) defective and (ii) non-defective. Although these approaches have obtained promising results, there exists two significant challenges in this research field: (i) removing irrelevant and redundant information from designing structures ; (ii) reducing the impact of skewed data distribution on learning models. In this paper, we aim to address these two issues by firstly applying kernel PCA to extract essential information from designing features and secondly proposing a deep neural network model which investigates the non-linear relationship among features. In order to mitigate the class imbalance, we apply a weighted loss function combined with a bootstrapping method to handle batch training mechanism of our model. We conducted some experiments to assess the performance of our proposed approach over NASA (with 10 projects) and PROMISE (with 34 projects) datasets. In order to leverage the efficiency of kernel PCA technique in software defect prediction, we compared it to some traditional feature selection approaches over a high-dimensional dataset ECLIPSE. The empirical results showed that our proposed method has outperformed these other state-of-the-art models by effectively predicting defective source files.