NatGen: Generative pre-training by "Naturalizing" source code.
Saikat ChakrabortyToufique AhmedYangruibo DingPremkumar T. DevanbuBaishakhi RayPublished in: CoRR (2022)
Keyphrases
- source code
- open source
- software systems
- software maintenance
- software projects
- open source software
- high level
- execution traces
- static analysis
- plagiarism detection
- source files
- legacy systems
- software evolution
- bug reports
- source code metrics
- change impact analysis
- free software
- symbolic execution
- authorship attribution
- maintenance activities
- program understanding
- program comprehension
- object oriented systems
- software engineers
- mailing lists
- visual basic
- website
- software engineering
- open source projects
- code reuse
- software repositories