Deep learning similarities from different representations of source code.
Michele TufanoCody WatsonGabriele BavotaMassimiliano Di PentaMartin WhiteDenys PoshyvanykPublished in: MSR (2018)
Keyphrases
- source code
- deep learning
- unsupervised feature learning
- deep belief networks
- open source
- software systems
- unsupervised learning
- machine learning
- plagiarism detection
- software projects
- software maintenance
- mental models
- program understanding
- software evolution
- weakly supervised
- similarity measure
- software repositories
- co occurrence
- high level
- named entities
- maximum likelihood
- free software
- source files
- legacy software