KNOD: Domain Knowledge Distilled Tree Decoder for Automated Program Repair
Automated Program Repair (APR) improves software reliability by generating patches for a buggy program automatically. Recent APR techniques leverage deep learning (DL) to build models to learn to generate patches from existing patches and code corpora. While promising, DL-based APR techniques suffer from the abundant syntactically or semantically incorrect patches in the patch space. These patches often disobey the syntactic and semantic domain knowledge of source code and thus cannot be the correct patches to fix a bug.
We propose a DL-based APR approach KNOD, which incorporates domain knowledge to guide patch generation in a direct and comprehensive way. KNOD has two major novelties, including (1) a novel three-stage tree decoder, which directly generates Abstract Syntax Trees of patched code according to the inherent tree structure, and (2) a novel domain-rule distillation, which leverages syntactic and semantic rules and teacher-student distributions to explicitly inject the domain knowledge into the decoding procedure during both the training and inference phases.
We evaluate KNOD on three widely-used benchmarks. KNOD fixes 72 bugs on the Defects4J v1.2, 25 bugs on the QuixBugs, and 50 bugs on the additional Defects4J v2.0 benchmarks, outperforming all existing APR tools.
Thu 18 MayDisplayed time zone: Hobart change
11:00 - 12:30 | Program repair techniques and applicationsTechnical Track / Journal-First Papers / DEMO - Demonstrations at Meeting Room 104 Chair(s): Xuan-Bach D. Le University of Melbourne | ||
11:00 15mTalk | Better Automatic Program Repair by Using Bug Reports and Tests Together Technical Track Pre-print | ||
11:15 15mTalk | CCTEST: Testing and Repairing Code Completion Systems Technical Track Li Zongjie , Chaozheng Wang Harbin Institute of Technology, Zhibo Liu Hong Kong University of Science and Technology, Haoxuan Wang EPFL, Dong Chen HKUST, Shuai Wang Hong Kong University of Science and Technology, Cuiyun Gao Harbin Institute of Technology | ||
11:30 7mTalk | A Controlled Experiment of Different Code Representations for Learning-Based Program Repair Journal-First Papers Marjane Namavar University of British Columbia, Noor Nashid University of British Columbia, Ali Mesbah University of British Columbia (UBC) Link to publication Pre-print | ||
11:37 7mTalk | Patching Locking Bugs Statically with Crayons Journal-First Papers Juan Alfredo Cruz-Carlon IT University of Copenhagen, Mahsa Varshosaz IT University of Copenhagen, Denmark, Claire Le Goues Carnegie Mellon University, Andrzej Wąsowski IT University of Copenhagen, Denmark | ||
11:45 15mTalk | KNOD: Domain Knowledge Distilled Tree Decoder for Automated Program Repair Technical Track Nan Jiang Purdue University, Thibaud Lutellier University of Alberta, Yiling Lou Fudan University, Lin Tan Purdue University, Dan Goldwasser Purdue University, Xiangyu Zhang Purdue University Pre-print | ||
12:00 15mTalk | Rete: Learning Namespace Representation for Program Repair Technical Track Nikhil Parasaram University College London, Earl T. Barr University College London, Sergey Mechtaev University College London Link to publication Pre-print | ||
12:15 7mTalk | Cerberus: a Program Repair Framework DEMO - Demonstrations Ridwan Salihin Shariffdeen National University of Singapore, Martin Mirchev National University of Singapore, Yannic Noller National University of Singapore, Abhik Roychoudhury National University of Singapore | ||
12:22 7mTalk | Predicting Patch Correctness Based on the Similarity of Failing Test Cases Journal-First Papers Haoye Tian University of Luxembourg, Yinghua LI University of Luxembourg, Weiguo PIAN University of Luxembourg, Abdoul Kader Kaboré SnT, University of Luxembourg, Kui Liu Huawei Software Engineering Application Technology Lab, Andrew Habib SnT, University of Luxembourg, Jacques Klein University of Luxembourg, Tegawendé F. Bissyandé SnT, University of Luxembourg | ||