Filters








152,181 Hits in 7.0 sec

TypeWriter: Neural Type Prediction with Search-based Validation [article]

Michael Pradel, Georgios Gousios, Jason Liu, Satish Chandra
2020 arXiv   pre-print
TypeWriter's predictor learns to infer the return and argument types for functions from partially annotated code bases by combining the natural language properties of code with programming language-level  ...  information.  ...  The model hence learns to predict "unknown" whenever none of the types in the vocabulary fit the given context information.  ... 
arXiv:1912.03768v2 fatcat:uiy6mvwfuvfvni5w23fjwy6jcq

Repository-Level Prompt Generation for Large Language Models of Code [article]

Disha Shrivastava, Hugo Larochelle, Daniel Tarlow
2022 arXiv   pre-print
In this work, we propose a framework called Repo-Level Prompt Generator that learns to generate example-specific prompts using a set of rules.  ...  These rules take context from the entire repository, thereby incorporating both the structure of the repository and the context from other relevant files (e.g. imports, parent class files).  ...  We would also like to extend our thanks to Breandan Considine for help in crawling the Google Code data archives; Justine Gehring, Avinash Bhat and Breandan Considine for helping with resources for running  ... 
arXiv:2206.12839v1 fatcat:3j5zjqh3vvhrzisd6dxrbddkhm

Review4Repair: Code Review Aided Automatic Program Repairing [article]

Faria Huq, Masum Hasan, Mahim Anzum Haque Pantho, Sazan Mahbub, Anindya Iqbal, Toufique Ahmed
2020 arXiv   pre-print
Context: Learning-based automatic program repair techniques are showing promise to provide quality fix suggestions for detected bugs in the source code of the software.  ...  However, none of the learning-based tools has utilized the review comments to fix programming bugs to the best of our knowledge.  ...  Model learns to predict code change ( f ), from code before change (C d ), defect location, and code review comment (R).  ... 
arXiv:2010.01544v2 fatcat:47yu7qdlebgavfo64r5dtbidqu

Towards Learning Generalizable Code Embeddings using Task-agnostic Graph Convolutional Networks

Zishuo Ding, Heng Li, Weiyi Shang, Tse-Hsun (Peter) Chen
2022 ACM Transactions on Software Engineering and Methodology  
Our findings suggest that future research and practice may consider using graph-based deep learning methods to capture the structural information of the source code for SE tasks.  ...  To evaluate the effectiveness of GraphCodeVec , we consider three downstream benchmark tasks (i.e., code comment generation, code authorship identification, and code clones detection) that are used in  ...  Diferent from CBOW which utilizes the context words to predict the target one, Skip-gram model tries to predict the surrounding context words given the target word.  ... 
doi:10.1145/3542944 fatcat:lthri6wilzgzdplgvaqgszl4au

Learning to Update Natural Language Comments Based on Code Changes [article]

Sheena Panthaplackel, Pengyu Nie, Milos Gligoric, Junyi Jessy Li, Raymond J. Mooney
2020 arXiv   pre-print
We propose an approach that learns to correlate changes across two distinct language representations, to generate a sequence of edits that are applied to the existing comment to reflect the source code  ...  We formulate the novel task of automatically updating an existing natural language comment based on changes in the body of code it accompanies.  ...  Namely, our model is trained to generate a sequence of edit actions, which are to be applied to the existing comment, by conditioning on learned representations of the code edits and existing comment.  ... 
arXiv:2004.12169v2 fatcat:bn2zlb62njhf3ap7dp2himebyy

A Survey of Automatic Software Vulnerability Detection, Program Repair, and Defect Prediction Techniques

Zhidong Shen, Si Chen, Luigi Coppolino
2020 Security and Communication Networks  
At the same time, we point out some problems of these research methods, give corresponding solutions, and finally look forward to the application prospect of deep learning technology in automated software  ...  vulnerability detection, automated program repair, and automated defect prediction.  ...  In the previous forecasting model, only the program code data was focused on, and the program code comment information was rarely paid attention to.  ... 
doi:10.1155/2020/8858010 fatcat:obeiw4p7afan5m24ydmdkmyhbm

Guest Editorial: Knowledge Discovery for Software Development (KDSD)

2020 IET Software  
Acknowledgments We are grateful to The Journal of IET Software Editor-in-Chief and the Editorial Office for their support throughout the editorial process.  ...  We would like to oblige all participants, Committee members, and the reviewers for this Special Issue of IET Software, for their dedication and hard work.  ...  to integrate domain of Machine Learning (ML), Statistical learning techniques, and Information Retrieval techniques.  ... 
doi:10.1049/iet-sen.2020.0166 fatcat:ixaksxisz5g55ml2arl2uvgzu4

A Survey of Machine Learning for Big Code and Naturalness

Miltiadis Allamanis, Earl T. Barr, Premkumar Devanbu, Charles Sutton
2018 ACM Computing Surveys  
Research at the intersection of machine learning, programming languages, and software engineering has recently taken important steps in proposing learnable probabilistic models of source code that exploit  ...  We present a taxonomy based on the underlying design principles of each model and use it to navigate the literature.  ...  to model context information.  ... 
doi:10.1145/3212695 fatcat:iuuocyctg5adjmobhc2zw23rfu

InCoder: A Generative Model for Code Infilling and Synthesis [article]

Daniel Fried, Armen Aghajanyan, Jessy Lin, Sida Wang, Eric Wallace, Freda Shi, Ruiqi Zhong, Wen-tau Yih, Luke Zettlemoyer, Mike Lewis
2022 arXiv   pre-print
Our model is the first generative model that is able to directly perform zero-shot code infilling, which we evaluate on challenging tasks such as type inference, comment generation, and variable re-naming  ...  bidirectional context.  ...  Acknowledgments We thank the Code Clippy team for open sourcing their data acquisition and deduplication code, 19 which we used as a starting point for our corpus collection.  ... 
arXiv:2204.05999v2 fatcat:tfefjg5ws5g3vm2iuc625kk2ey

Hippocampal Representational Organization and Spatial Context

S.J.Y. Mizumori, K.E. Ragozzino, B.G. Cooper, S. Leutgeb
1999 Hippocampus  
The specific contribution of the hilar/CA3 region is suggested to be to compare the expected spatial context with that currently being experienced, then relay discrepancies to CA1.  ...  In this way, hippocampus helps to distinguish temporally one spatial context from another, thereby contributing to episodic memories. Hippocampus 1999;9:444-451.  ...  Acknowledgments We thank James Canfield and Wayne Pratt for comments regarding this manuscript.  ... 
doi:10.1002/(sici)1098-1063(1999)9:4<444::aid-hipo10>3.0.co;2-z pmid:10495025 fatcat:ug7lvg5zwbdttdvovtabufu5jm

Hippocampal Representational Organization and Spatial Context

S.J.Y. Mizumori, K.E. Ragozzino, B.G. Cooper, S. Leutgeb
1999 Hippocampus  
The specific contribution of the hilar/CA3 region is suggested to be to compare the expected spatial context with that currently being experienced, then relay discrepancies to CA1.  ...  In this way, hippocampus helps to distinguish temporally one spatial context from another, thereby contributing to episodic memories. Hippocampus 1999;9:444-451.  ...  Acknowledgments We thank James Canfield and Wayne Pratt for comments regarding this manuscript.  ... 
doi:10.1002/(sici)1098-1063(1999)9:4<444::aid-hipo10>3.3.co;2-q pmid:10495025 fatcat:xkjjsl3irfg3nbru3hcnlleymu

A Survey of Machine Learning for Big Code and Naturalness [article]

Miltiadis Allamanis, Earl T. Barr, Premkumar Devanbu, Charles Sutton
2018 arXiv   pre-print
Research at the intersection of machine learning, programming languages, and software engineering has recently taken important steps in proposing learnable probabilistic models of source code that exploit  ...  We present a taxonomy based on the underlying design principles of each model and use it to navigate the literature.  ...  to model context information.  ... 
arXiv:1709.06182v2 fatcat:hbvgyonqsjgq3nqwji6jf3aybe

Less Training, More Repairing Please: Revisiting Automated Program Repair via Zero-shot Learning [article]

Chunqiu Steven Xia, Lingming Zhang
2022 arXiv   pre-print
Our main insight is instead of modeling what a repair edit should look like, we can directly predict what the correct code is based on the context information.  ...  Therefore, in this paper, we aim to revisit the learning-based APR problem, and propose AlphaRepair, to leverage zero-shot learning directly using large pre-trained code models for APR.  ...  We build AlphaRepair using Code-BERT and design inputs to make use of the pre-training objective of CodeBERT to directly generate fix lines from the surrounding context.  ... 
arXiv:2207.08281v2 fatcat:5eucjvil3vdrhollpdoa6l7hzm

Improving automatic source code summarization via deep reinforcement learning

Yao Wan, Zhou Zhao, Min Yang, Guandong Xu, Haochao Ying, Jian Wu, Philip S. Yu
2018 Proceedings of the 33rd ACM/IEEE International Conference on Automated Software Engineering - ASE 2018  
The actor network provides the confidence of predicting the next word according to current state.  ...  trained to predict the next word by maximizing the likelihood of next groundtruth word with previous ground-truth word given.  ...  LSTM model [18] to represent the structure of code. We also use another LSTM model [42] to represent the sequential information of code.  ... 
doi:10.1145/3238147.3238206 dblp:conf/kbse/WanZYXY0Y18 fatcat:eknsug7narft5dre5ncenn2nvm

Dipole

Fenglong Ma, Radha Chitta, Jing Zhou, Quanzeng You, Tong Sun, Jing Gao
2017 Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '17  
Dipole also allows us to interpret the learned medical code representations which are confirmed positively by medical experts.  ...  To address these issues, we propose Dipole, an end-to-end, simple and robust model for predicting patients' future health information.  ...  ACKNOWLEDGMENTS e authors would like to thank the anonymous referees for their valuable comments and helpful suggestions. is work is supported in part by the US National Science Foundation under grants  ... 
doi:10.1145/3097983.3098088 dblp:conf/kdd/MaCZYSG17 fatcat:6sth7fjv7bcwlhsfzl4zu6oz6u
« Previous Showing results 1 — 15 out of 152,181 results