Hong Jin Kang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Hong Jin Kang has a total of 48 co-authors. Here are the top co-authors by number of shared publications:
- David Lo
Singapore Management University
12 shared publications - Ratnadira Widyasari
Singapore Management University, Singapore, Singapore
5 shared publications - Eng Lieh Ouh
Singapore Management University, Singapore
4 shared publications - Jieke Shi
Singapore Management University, Singapore, Singapore
4 shared publications - Lwin Khin Shar
Singapore Management University, Singapore, Singapore
4 shared publications - Zhou Yang
University of Alberta, Edmonton, Canada
4 shared publications
Latest Publications
- Open Access
- 7 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 298-309
Maximal Format-Free Data Repair
- 7 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 298-309
Open AccessMaximal Format-Free Data Repair
- Open Access
- 23 May 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Software Engineering and Methodology
Mapping NVD Records to Their Vulnerability-fixing Commits: How Hard is It?
Mapping National Vulnerability Database (NVD) records to vulnerability-fixing commits (VFCs) is crucial for vulnerability analysis but challenging due to sparse explicit links in NVD references. This study explores this mapping's feasibility through an empirical approach. Manual analysis of NVD references showed Git references enable over 86% success, while non-Git references achieve under 14%. Using these findings, we built an automated pipeline extracting 31,942 VFCs from 20,360 NVD records (8.7% of 235,341) with 87% precision, mainly from Git references. To fill gaps, we mined six external security databases, yielding 29,254 VFCs for 18,985 records (8.1%) at 88.4% precision, and GitHub repositories, adding 3,686 VFCs for 2,795 records (1.2%) at 73% precision. Combining these, we mapped 26,710 unique records (11.3% coverage) from 7,634 projects, with overlap between NVD and external databases, plus unique GitHub contributions. Despite success with Git references, 88.7% of records remain unmapped, highlighting the difficulty without Git links. This study offers insights for enhancing vulnerability datasets and guiding future automated security research.
- 23 May 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Software Engineering and Methodology
Open AccessMapping NVD Records to Their Vulnerability-fixing Commits: How Hard is It?
Mapping National Vulnerability Database (NVD) records to vulnerability-fixing commits (VFCs) is crucial for vulnerability analysis but challenging due to sparse explicit links in NVD references. This study explores this mapping's feasibility through an empirical approach. Manual analysis of NVD references showed Git references enable over 86% success, while non-Git references achieve under 14%. Using these findings, we built an automated pipeline extracting 31,942 VFCs from 20,360 NVD records (8.7% of 235,341) with 87% precision, mainly from Git references. To fill gaps, we mined six external security databases, yielding 29,254 VFCs for 18,985 records (8.1%) at 88.4% precision, and GitHub repositories, adding 3,686 VFCs for 2,795 records (1.2%) at 73% precision. Combining these, we mapped 26,710 unique records (11.3% coverage) from 7,634 projects, with overlap between NVD and external databases, plus unique GitHub contributions. Despite success with Git references, 88.7% of records remain unmapped, highlighting the difficulty without Git links. This study offers insights for enhancing vulnerability datasets and guiding future automated security research.
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 2466-2478
WhyFlow: Interrogative Debugger for Sensemaking Taint Analysis
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 2466-2478
Open AccessWhyFlow: Interrogative Debugger for Sensemaking Taint Analysis
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 580-591
Think Like Human Developers: Harnessing Community Knowledge for Structured Code Reasoning
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 580-591
Open AccessThink Like Human Developers: Harnessing Community Knowledge for Structured Code Reasoning
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 2161-2173
Back to the Basics: Rethinking Issue-Commit Linking with LLM-Assisted Retrieval
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 2161-2173
Open AccessBack to the Basics: Rethinking Issue-Commit Linking with LLM-Assisted Retrieval
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1030-1042
Let the Trial Begin: A Mock-Court Approach to Vulnerability Detection using LLM-Based Agents
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 1030-1042
Open AccessLet the Trial Begin: A Mock-Court Approach to Vulnerability Detection using LLM-Based Agents
- Open Access
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 186-190
Abductive Reasoning for Neurosymbolic Fault Localization
Fault localization is essential for debugging. Recently, Large Language Model (LLM)-based techniques have been proposed to identify buggy methods from failing test cases. The state-of-the-art approach, SoapFL, adopts a structured multi-phase procedure. Given a test failure, suspicious classes and methods are identified based on its analysis of the test failure. We conduct a funnel analysis of the multi-phase approach. While effective, our analysis highlights a limitation: the early selection of a single hypothesis can cause useful alternatives to be discarded. Motivated by how developers consider multiple competing hypotheses of the root cause of a bug as they analyze the evidence, we prototype HypoDeduce, which applies abductive reasoning. HypoDeduce generates and evaluates competing hypotheses, integrates evidence across classes and methods, and then identifies suspicious methods. On Defects4J, our prototype successfully localizes the most suspicious method for 48.6% of bugs, improving over SoapFL by 6%. These results suggest that structured hypothesis management enhances fault localization. We propose a research agenda for better supporting abductive analysis with more principled forms of reasoning supported by neurosymbolic approaches. In particular, we highlight the need for probabilistic reasoning, a combination of neural and symbolic methods for code navigation, as well as counterfactual analysis for suspicious code selection.
- 12 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 186-190
Open AccessAbductive Reasoning for Neurosymbolic Fault Localization
Fault localization is essential for debugging. Recently, Large Language Model (LLM)-based techniques have been proposed to identify buggy methods from failing test cases. The state-of-the-art approach, SoapFL, adopts a structured multi-phase procedure. Given a test failure, suspicious classes and methods are identified based on its analysis of the test failure. We conduct a funnel analysis of the multi-phase approach. While effective, our analysis highlights a limitation: the early selection of a single hypothesis can cause useful alternatives to be discarded. Motivated by how developers consider multiple competing hypotheses of the root cause of a bug as they analyze the evidence, we prototype HypoDeduce, which applies abductive reasoning. HypoDeduce generates and evaluates competing hypotheses, integrates evidence across classes and methods, and then identifies suspicious methods. On Defects4J, our prototype successfully localizes the most suspicious method for 48.6% of bugs, improving over SoapFL by 6%. These results suggest that structured hypothesis management enhances fault localization. We propose a research agenda for better supporting abductive analysis with more principled forms of reasoning supported by neurosymbolic approaches. In particular, we highlight the need for probabilistic reasoning, a combination of neural and symbolic methods for code navigation, as well as counterfactual analysis for suspicious code selection.
- Open Access
- 7 April 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Software Engineering and Methodology
Revisiting Vulnerability Patch Identification on Data in the Wild
Attacks can exploit zero-day or one-day vulnerabilities that are not publicly disclosed. To detect these vulnerabilities, security researchers monitor development activities in open-source repositories to identify unreported security patches. The sheer volume of commits makes this task infeasible to accomplish manually. Consequently, security patch detectors commonly trained and evaluated on security patches linked from vulnerability reports in the National Vulnerability Database (NVD). In this study, we assess the effectiveness of these detectors when applied in-the-wild. Our results show that models trained on NVD-derived data show substantially decreased performance, with decreases in F1-score of up to 90% when tested on in-the-wild security patches, rendering them impractical for real-world use. An analysis comparing security patches identified in-the-wild and commits linked from NVD reveals that they can be easily distinguished from each other. Security patches associated with NVD have different distribution of commit messages, vulnerability types, and composition of changes. These differences suggest that NVD may be unsuitable as the sole source of data for training models to detect security patches. We find that constructing a dataset that combines security patches from NVD data with a small subset of manually identified security patches can improve model robustness.
- 7 April 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Software Engineering and Methodology
Open AccessRevisiting Vulnerability Patch Identification on Data in the Wild
Attacks can exploit zero-day or one-day vulnerabilities that are not publicly disclosed. To detect these vulnerabilities, security researchers monitor development activities in open-source repositories to identify unreported security patches. The sheer volume of commits makes this task infeasible to accomplish manually. Consequently, security patch detectors commonly trained and evaluated on security patches linked from vulnerability reports in the National Vulnerability Database (NVD). In this study, we assess the effectiveness of these detectors when applied in-the-wild. Our results show that models trained on NVD-derived data show substantially decreased performance, with decreases in F1-score of up to 90% when tested on in-the-wild security patches, rendering them impractical for real-world use. An analysis comparing security patches identified in-the-wild and commits linked from NVD reveals that they can be easily distinguished from each other. Security patches associated with NVD have different distribution of commit messages, vulnerability types, and composition of changes. These differences suggest that NVD may be unsuitable as the sole source of data for training models to detect security patches. We find that constructing a dataset that combines security patches from NVD data with a small subset of manually identified security patches can improve model robustness.
- Open Access
- 23 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 204-214
Grey-Box Fuzzing in Constrained Ultra-Large Systems: Lessons for SE Community
- 23 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 204-214
Open AccessGrey-Box Fuzzing in Constrained Ultra-Large Systems: Lessons for SE Community
- Open Access
- 19 June 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Software Engineering
- Vol. 2 (FSE) , 647-667
DuoReduce: Bug Isolation for Multi-layer Extensible Compilation
In recent years, the MLIR framework has had explosive growth due to the need for extensible deep learning compilers for hardware accelerators. Such examples include Triton, CIRCT, and ONNX-MLIR. MLIR compilers introduce significant complexities in localizing bugs or inefficiencies because of their layered optimization and transformation process with compilation passes. While existing delta debugging techniques can be used to identify a minimum subset of IR code that reproduces a given bug symptom, their naive application to MLIR is time-consuming because real-world MLIR compilers usually involve a large number of compilation passes. Compiler developers must identify a minimized set of relevant compilation passes to reduce the footprint of MLIR compiler code to be inspected for a bug fix. We propose DuoReduce, a dual-dimensional reduction approach for MLIR bug localization. DuoReduce leverages three key ideas in tandem to design an efficient MLIR delta debugger. First, DuoReduce reduces compiler passes that are irrelevant to the bug by identifying ordering dependencies among the different compilation passes. Second, DuoReduce uses MLIR-semantics-aware transformations to expedite IR code reduction. Finally, DuoReduce leverages cross-dependence between the IR code dimension and the compilation pass dimension by accounting for which IR code segments are related to which compilation passes to reduce unused passes. Experiments with three large-scale MLIR compiler projects find that DuoReduce outperforms syntax-aware reducers such as Perses and Vulcan in terms of IR code reduction by 31.6% and 21.5% respectively. If one uses these reducers by enumerating all possible compilation passes (on average 18 passes), it could take up to 145 hours. By identifying ordering dependencies among compilation passes, DuoReduce reduces this time to 9.5 minutes. By identifying which compilation passes are unused for compiling reduced IR code, DuoReduce reduces the number of passes by 14.6%. This translates to not needing to examine 281 lines of MLIR compiler code on average to fix the bugs. DuoReduce has the potential to significantly reduce debugging effort in MLIR compilers, which serves as the foundation for the current landscape of machine learning and hardware accelerators.
- 19 June 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Software Engineering
- Vol. 2 (FSE) , 647-667
Open AccessDuoReduce: Bug Isolation for Multi-layer Extensible Compilation
In recent years, the MLIR framework has had explosive growth due to the need for extensible deep learning compilers for hardware accelerators. Such examples include Triton, CIRCT, and ONNX-MLIR. MLIR compilers introduce significant complexities in localizing bugs or inefficiencies because of their layered optimization and transformation process with compilation passes. While existing delta debugging techniques can be used to identify a minimum subset of IR code that reproduces a given bug symptom, their naive application to MLIR is time-consuming because real-world MLIR compilers usually involve a large number of compilation passes. Compiler developers must identify a minimized set of relevant compilation passes to reduce the footprint of MLIR compiler code to be inspected for a bug fix. We propose DuoReduce, a dual-dimensional reduction approach for MLIR bug localization. DuoReduce leverages three key ideas in tandem to design an efficient MLIR delta debugger. First, DuoReduce reduces compiler passes that are irrelevant to the bug by identifying ordering dependencies among the different compilation passes. Second, DuoReduce uses MLIR-semantics-aware transformations to expedite IR code reduction. Finally, DuoReduce leverages cross-dependence between the IR code dimension and the compilation pass dimension by accounting for which IR code segments are related to which compilation passes to reduce unused passes. Experiments with three large-scale MLIR compiler projects find that DuoReduce outperforms syntax-aware reducers such as Perses and Vulcan in terms of IR code reduction by 31.6% and 21.5% respectively. If one uses these reducers by enumerating all possible compilation passes (on average 18 passes), it could take up to 145 hours. By identifying ordering dependencies among compilation passes, DuoReduce reduces this time to 9.5 minutes. By identifying which compilation passes are unused for compiling reduced IR code, DuoReduce reduces the number of passes by 14.6%. This translates to not needing to examine 281 lines of MLIR compiler code on average to fix the bugs. DuoReduce has the potential to significantly reduce debugging effort in MLIR compilers, which serves as the foundation for the current landscape of machine learning and hardware accelerators.