CONCORD: Clone-Aware Contrastive Learning for Source Code
ACM SIGSOFT Distinguished Paper
Deep Learning (DL) models to analyze source code have shown immense promise during the past few years.
More recently, self-supervised pre-training has gained traction for learning generic code representations valuable for many downstream SE tasks, such as clone and bug detection.
While previous work successfully learned from different code abstractions (e.g., token, AST, graph), we argue that it is also essential to factor in how developers code day-to-day for learning general-purpose representation. On the one hand, human developers tend to write repetitive programs referencing existing code snippets from the current codebase or online resources (e.g., Stack Overflow website) rather than implementing functions from scratch; such behaviors result in a vast number of code clones. In contrast, a deviant clone by mistake might trigger malicious program behaviors.
Thus, as a proxy to incorporate developers' coding behavior into the pre-training scheme, we propose to include code clones and their deviants. In particular, we propose CONCORD, a self-supervised pre-training strategy to place benign clones closer in the representation space while moving deviants further apart. We show that CONCORD's clone-aware pre-training drastically reduces the need for expensive pre-training resources while improving the performance of downstream SE tasks. We also empirically demonstrate that CONCORD can improve existing pre-trained models to learn better representations that consequently become more efficient in both identifying semantically equivalent programs and differentiating buggy from non-buggy code.
Tue 18 JulDisplayed time zone: Pacific Time (US & Canada) change
13:30 - 15:00
|API2Vec: Learning Representations of API Sequences for Malware Detection
Lei Cui Zhongguancun Laboratory, Jiancong Cui University of Chinese Academy of Sciences; Institute of Information Engineering at Chinese Academy of Sciences, Yuede Ji University of North Texas, Zhiyu Hao Zhongguancun Laboratory, Lun Li Institute of Information Engineering at Chinese Academy of Sciences, Zhenquan Ding Institute of Information Engineering at Chinese Academy of SciencesDOI
|CONCORD: Clone-Aware Contrastive Learning for Source CodeACM SIGSOFT Distinguished Paper
Yangruibo Ding Columbia University, Saikat Chakraborty Microsoft Research, Luca Buratti IBM Research, Saurabh Pujar IBM, Alessandro Morari IBM Research, Gail Kaiser Columbia University, Baishakhi Ray Columbia UniversityDOI
|Type Batched Program Reduction
|CodeGrid: A Grid Representation of Code
Abdoul Kader Kaboré University of Luxembourg, Earl T. Barr University College London; Google DeepMind, Jacques Klein University of Luxembourg, Tegawendé F. Bissyandé University of LuxembourgDOI
|Guided Retraining to Enhance the Detection of Difficult Android Malware
Nadia Daoudi University of Luxembourg, Kevin Allix CentraleSupélec, Tegawendé F. Bissyandé University of Luxembourg, Jacques Klein University of LuxembourgDOI
|Automatically Reproducing Android Bug Reports using Natural Language Processing and Reinforcement Learning
Zhaoxu Zhang University of Southern California, Robert Winn University of Southern California, Yu Zhao University of Central Missouri, Tingting Yu University of Cincinnati, William G.J. Halfond University of Southern CaliforniaDOI