Patanamon Thongtanunam

Hong Yi Lin, Patanamon Thongtanunam, Christoph Treude et al.

Modern code review is a ubiquitous software quality assurance process aimed at identifying potential issues within newly written code. Despite its effectiveness, the process demands large amounts of effort from the human reviewers involved. To help alleviate this workload, researchers have trained deep learning models to imitate human reviewers in providing natural language code reviews. Formally, this task is known as code review comment generation. Prior work has demonstrated improvements in this task by leveraging machine learning techniques and neural models, such as transfer learning and the transformer architecture. However, the quality of the model generated reviews remain sub-optimal due to the quality of the open-source code review data used in model training. This is in part due to the data obtained from open-source projects where code reviews are conducted in a public forum, and reviewers possess varying levels of software development experience, potentially affecting the quality of their feedback. To accommodate for this variation, we propose a suite of experience-aware training methods that utilise the reviewers' past authoring and reviewing experiences as signals for review quality. Specifically, we propose experience-aware loss functions (ELF), which use the reviewers' authoring and reviewing ownership of a project as weights in the model's loss function. Through this method, experienced reviewers' code reviews yield larger influence over the model's behaviour. Compared to the SOTA model, ELF was able to generate higher quality reviews in terms of accuracy, informativeness, and comment types generated. The key contribution of this work is the demonstration of how traditional software engineering concepts such as reviewer experience can be integrated into the design of AI-based automated code review models.

Peerachai Banyongrakkul, Mansooreh Zahedi, Patanamon Thongtanunam et al.

Pre-trained models (PTMs) have gained widespread popularity and achieved remarkable success across various fields, driven by their groundbreaking performance and easy accessibility through hosting providers. However, the challenges faced by downstream developers in reusing PTMs in software systems are less explored. To bridge this knowledge gap, we qualitatively created and analyzed a dataset of 840 PTM-related issue reports from 31 OSS GitHub projects. We systematically developed a comprehensive taxonomy of PTM-related challenges that developers face in downstream projects. Our study identifies seven key categories of challenges that downstream developers face in reusing PTMs, such as model usage, model performance, and output quality. We also compared our findings with existing taxonomies. Additionally, we conducted a resolution time analysis and, based on statistical tests, found that PTM-related issues take significantly longer to be resolved than issues unrelated to PTMs, with significant variation across challenge categories. We discuss the implications of our findings for practitioners and possibilities for future research.

Wachiraphan Charoenwet, Kla Tantithamthavorn, Patanamon Thongtanunam et al.

Secure code review is critical at the pre-commit stage, where vulnerabilities must be caught early under tight latency and limited-context constraints. Existing SAST-based checks are noisy and often miss immature, context-dependent vulnerabilities, while standalone Large Language Models (LLMs) are constrained by context windows and lack explicit tool use. Agentic AI, which combine LLMs with autonomous decision-making, tool invocation, and code navigation, offer a promising alternative, but their effectiveness for pre-commit secure code review is not yet well understood. In this work, we introduce AgenticSCR, an agentic AI for secure code review for detecting immature vulnerabilities during the pre-commit stage, augmented by security-focused semantic memories. Using our own curated benchmark of immature vulnerabilities, tailored to the pre-commit secure code review, we empirically evaluate how accurate is our AgenticSCR for localizing, detecting, and explaining immature vulnerabilities. Our results show that AgenticSCR achieves at least 153% relatively higher percentage of correct code review comments than the static LLM-based baseline, and also substantially surpasses SAST tools. Moreover, AgenticSCR generates more correct comments in four out of five vulnerability types, consistently and significantly outperforming all other baselines. These findings highlight the importance of Agentic Secure Code Review, paving the way towards an emerging research area of immature vulnerability detection.

Supatsara Wattanakriengkrai, Patanamon Thongtanunam, Chakkrit Tantithamthavorn et al.

Defect prediction models are proposed to help a team prioritize source code areas files that need Software QualityAssurance (SQA) based on the likelihood of having defects. However, developers may waste their unnecessary effort on the whole filewhile only a small fraction of its source code lines are defective. Indeed, we find that as little as 1%-3% of lines of a file are defective. Hence, in this work, we propose a novel framework (called LINE-DP) to identify defective lines using a model-agnostic technique, i.e., an Explainable AI technique that provides information why the model makes such a prediction. Broadly speaking, our LINE-DP first builds a file-level defect model using code token features. Then, our LINE-DP uses a state-of-the-art model-agnostic technique (i.e.,LIME) to identify risky tokens, i.e., code tokens that lead the file-level defect model to predict that the file will be defective. Then, the lines that contain risky tokens are predicted as defective lines. Through a case study of 32 releases of nine Java open source systems, our evaluation results show that our LINE-DP achieves an average recall of 0.61, a false alarm rate of 0.47, a top 20%LOC recall of0.27, and an initial false alarm of 16, which are statistically better than six baseline approaches. Our evaluation shows that our LINE-DP requires an average computation time of 10 seconds including model construction and defective line identification time. In addition, we find that 63% of defective lines that can be identified by our LINE-DP are related to common defects (e.g., argument change, condition change). These results suggest that our LINE-DP can effectively identify defective lines that contain common defectswhile requiring a smaller amount of inspection effort and a manageable computation cost.

Michael R. Lyu, Baishakhi Ray, Abhik Roychoudhury et al.

Automatic programming has seen increasing popularity due to the emergence of tools like GitHub Copilot which rely on Large Language Models (LLMs). At the same time, automatically generated code faces challenges during deployment due to concerns around quality and trust. In this article, we study automated coding in a general sense and study the concerns around code quality, security and related issues of programmer responsibility. These are key issues for organizations while deciding on the usage of automatically generated code. We discuss how advances in software engineering such as program repair and analysis can enable automatic programming. We conclude with a forward looking view, focusing on the programming environment of the near future, where programmers may need to switch to different roles to fully utilize the power of automatic programming. Automated repair of automatically generated programs from LLMs, can help produce higher assurance code from LLMs, along with evidence of assurance

Wannita Takerngsaksiri, Jirat Pasuksmit, Patanamon Thongtanunam et al.

Recently, Large Language Models (LLMs)-based multi-agent paradigms for software engineering are introduced to automatically resolve software development tasks (e.g., from a given issue to source code). However, existing work is evaluated based on historical benchmark datasets, rarely considers human feedback at each stage of the automated software development process, and has not been deployed in practice. In this paper, we introduce a Human-in-the-loop LLM-based Agents framework (HULA) for software development that allows software engineers to refine and guide LLMs when generating coding plans and source code for a given task. We design, implement, and deploy the HULA framework into Atlassian JIRA for internal uses. Through a multi-stage evaluation of the HULA framework, Atlassian software engineers perceive that HULA can minimize the overall development time and effort, especially in initiating a coding plan and writing code for straightforward tasks. On the other hand, challenges around code quality remain a concern in some cases. We draw lessons learned and discuss opportunities for future work, which will pave the way for the advancement of LLM-based agents in software development.

Yang Hong, Chakkrit Tantithamthavorn, Jirat Pasuksmit et al.

Continuous Integration (CI) build failures could significantly impact the software development process and teams, such as delaying the release of new features and reducing developers' productivity. In this work, we report on an empirical study that investigates CI build failures throughout product development at Atlassian. Our quantitative analysis found that the repository dimension is the key factor influencing CI build failures. In addition, our qualitative survey revealed that Atlassian developers perceive CI build failures as challenging issues in practice. Furthermore, we found that the CI build prediction can not only provide proactive insight into CI build failures but also facilitate the team's decision-making. Our study sheds light on the challenges and expectations involved in integrating CI build prediction tools into the Bitbucket environment, providing valuable insights for enhancing CI processes.

Hong Yi Lin, Chunhua Liu, Haoyu Gao et al.

State-of-the-art large language models (LLMs) have demonstrated impressive code generation capabilities but struggle with real-world software engineering tasks, such as revising source code to address code reviews, hindering their practical use. Code review comments are often implicit, ambiguous, and colloquial, requiring models to grasp both code and human intent. This challenge calls for evaluating large language models' ability to bridge both technical and conversational contexts. While existing work has employed the automated code refinement (ACR) task to resolve these comments, current evaluation methods fall short, relying on text matching metrics that provide limited insight into model failures and remain susceptible to training data contamination. To address these limitations, we introduce a novel evaluation benchmark, $\textbf{CodeReviewQA}$ that enables us to conduct fine-grained assessment of model capabilities and mitigate data contamination risks. In CodeReviewQA, we decompose the generation task of code refinement into $\textbf{three essential reasoning steps}$: $\textit{change type recognition}$ (CTR), $\textit{change localisation}$ (CL), and $\textit{solution identification}$ (SI). Each step is reformulated as multiple-choice questions with varied difficulty levels, enabling precise assessment of model capabilities, while mitigating data contamination risks. Our comprehensive evaluation spans 72 recently released large language models on $\textbf{900 manually curated, high-quality examples}$ across nine programming languages. Our results show that CodeReviewQA is able to expose specific model weaknesses in code review comprehension, disentangled from their generative automated code refinement results.

Huy Nguyen, Christoph Treude, Patanamon Thongtanunam

Automated program comprehension underpins many software engineering tasks, from code summarisation to clone detection. Recent deep learning models achieve strong results but typically rely on source code alone, overlooking contextual information such as version history or structural relationships. This limits their ability to capture how code evolves and operates. We conduct an empirical study on how enriching code representations with such contextual signals affects neural model performance on key comprehension tasks. Two downstream tasks, code clone detection and code summarisation, are evaluated using SeSaMe (1,679 Java methods) and CodeSearchNet (63,259 methods). Five representative models (CodeBERT, GraphCodeBERT, CodeT5, PLBART, ASTNN) are fine-tuned under code-only and context-augmented settings. Results show that context generally improves performance: version history consistently boosts clone detection (e.g., CodeT5 +15.92% F1) and summarisation (e.g., GraphCodeBERT +5.56% METEOR), while call-graph effects vary by model and task. Combining multiple contexts yields further gains (up to +21.48% macro-F1). Human evaluation on 100 Java snippets confirms that context-augmented summaries are significantly preferred for Accuracy and Content Adequacy (p <= 0.026; |delta| up to 0.55). These findings highlight the potential of contextual signals to enhance code comprehension and open new directions for optimising contextual encoding in neural SE models.

Chunhua Liu, Hong Yi Lin, Patanamon Thongtanunam

Language models have shown strong capabilities across a wide range of tasks in software engineering, such as code generation, yet they suffer from hallucinations. While hallucinations have been studied independently in natural language and code generation, their occurrence in tasks involving code changes which have a structurally complex and context-dependent format of code remains largely unexplored. This paper presents the first comprehensive analysis of hallucinations in two critical tasks involving code change to natural language generation: commit message generation and code review comment generation. We quantify the prevalence of hallucinations in recent language models and explore a range of metric-based approaches to automatically detect them. Our findings reveal that approximately 50\% of generated code reviews and 20\% of generated commit messages contain hallucinations. Whilst commonly used metrics are weak detectors on their own, combining multiple metrics substantially improves performance. Notably, model confidence and feature attribution metrics effectively contribute to hallucination detection, showing promise for inference-time detection.\footnote{All code and data will be released upon acceptance.

Kyi Shin Khant, Hong Yi Lin, Patanamon Thongtanunam

Learning-based techniques, especially advanced pre-trained models for code have demonstrated capabilities in code understanding and generation, solving diverse software engineering (SE) tasks. Despite the promising results, current training approaches may not fully optimize model performance, as they typically involve learning from randomly shuffled training data. Recent work shows that Curriculum Learning (CL) can improve performance on code-related tasks through incremental learning based on the difficulty of synthetic code. Yet, the effectiveness of CL with conventional difficulty measures in SE tasks remains largely unexplored. In this study, we explore two conventional code metrics: code length and cyclomatic complexity to determine the difficulty levels. We investigate how the pre-trained code model (CodeT5) learns under CL, through the tasks of code clone detection and code summarization. Our empirical study on the CodeXGLUE benchmark showed contrasting results to prior studies, where the model exhibited signs of catastrophic forgetting and shortcut learning. Surprisingly, model performance saturates after only the first quartile of training, potentially indicating a limit in the model's representation capacity and/or the task's inherent difficulty. Future work should further explore various CL strategies with different code models across a wider range of SE tasks for a more holistic understanding.

Jirat Pasuksmit, Patanamon Thongtanunam, Shanika Karunasekera

Effort estimation is an integral part of activities planning in Agile iterative development. An Agile team estimates the effort of a task based on the available information which is usually conveyed through documentation. However, as documentation has a lower priority in Agile, little is known about how documentation effort can be optimized while achieving accurate estimation. Hence, to help practitioners achieve just-enough documentation for effort estimation, we investigated the different types of documented information that practitioners considered useful for effort estimation. We conducted a survey study with 121 Agile practitioners across 25 countries. Our survey results showed that (1) despite the lower priority of documentation in Agile practices, 98% of the respondents considered documented information moderately to extremely important when estimating effort, (2) 73% of them reported that they would re-estimate a task when the documented information was changed, and (3) functional requirements, user stories, definition of done, UI wireframes, acceptance criteria, and task dependencies were ranked as the most useful types of documented information for effort estimation. Nevertheless, many respondents reported that these useful types of documented information were occasionally changing or missing. Based on our study results, we provide recommendations for agile practitioners on how effort estimation can be improved by focusing on just-enough documentation.

Chun Yong Chong, Patanamon Thongtanunam, Chakkrit Tantithamthavorn

Code review is a widely-used practice in software development companies to identify defects. Hence, code review has been included in many software engineering curricula at universities worldwide. However, teaching code review is still a challenging task because the code review effectiveness depends on the code reading and analytical skills of a reviewer. While several studies have investigated the code reading techniques that students should use to find defects during code review, little has focused on a learning activity that involves analytical skills. Indeed, developing a code review checklist should stimulate students to develop their analytical skills to anticipate potential issues (i.e., software defects). Yet, it is unclear whether students can anticipate potential issues given their limited experience in software development (programming, testing, etc.). We perform a qualitative analysis to investigate whether students are capable of creating code review checklists, and if the checklists can be used to guide reviewers to find defects. In addition, we identify common mistakes that students make when developing a code review checklist. Our results show that while there are some misconceptions among students about the purpose of code review, students are able to anticipate potential defects and create a relatively good code review checklist. Hence, our results lead us to conclude that developing a code review checklist can be a part of the learning activities for code review in order to scaffold students' skills.

Anwar Alqaimi, Patanamon Thongtanunam, Christoph Treude

When lambda expressions were introduced to the Java programming language as part of the release of Java 8 in 2014, they were the language's first step into functional programming. Since lambda expressions are still relatively new, not all developers use or understand them. In this paper, we first present the results of an empirical study to determine how frequently developers of GitHub repositories make use of lambda expressions and how they are documented. We find that 11% of Java GitHub repositories use lambda expressions, and that only 6% of the lambda expressions are accompanied by source code comments. We then present a tool called LambdaDoc which can automatically detect lambda expressions in a Java repository and generate natural language documentation for them. Our evaluation of LambdaDoc with 23 professional developers shows that they perceive the generated documentation to be complete, concise, and expressive, while the majority of the documentation produced by our participants without tool support was inadequate. Our contribution builds an important step towards automatically generating documentation for functional programming constructs in an object-oriented language.

Shade Ruangwan, Patanamon Thongtanunam, Akinori Ihara et al.

Modern Code Review (MCR) plays a key role in software quality practices. In MCR process, a new patch (i.e., a set of code changes) is encouraged to be examined by reviewers in order to identify weaknesses in source code prior to an integration into main software repositories. To mitigate the risk of having future defects, prior work suggests that MCR should be performed with sufficient review participation. Indeed, recent work shows that a low number of participated reviewers is associated with poor software quality. However, there is a likely case that a new patch still suffers from poor review participation even though reviewers were invited. Hence, in this paper, we set out to investigate the factors that are associated with the participation decision of an invited reviewer. Through a case study of 230,090 patches spread across the Android, LibreOffice, OpenStack and Qt systems, we find that (1) 16%-66% of patches have at least one invited reviewer who did not respond to the review invitation; (2) human factors play an important role in predicting whether or not an invited reviewer will participate in a review; (3) a review participation rate of an invited reviewers and code authoring experience of an invited reviewer are highly associated with the participation decision of an invited reviewer. These results can help practitioners better understand about how human factors associate with the participation decision of reviewers and serve as guidelines for inviting reviewers, leading to a better inviting decision and a better reviewer participation.