turbolent/tokenizer
A Penn Treebank tokenizer
Lila is a dynamic language running on the JVM with efficient multiple and predicate method dispatch
Public repository record indexed from GitHub. Explore verified star velocity metrics, source code repositories, and curated developer tool directories across the GitHubRepo ecosystem.
A Penn Treebank tokenizer
A lemmatizer based on WordNet's morphological processing
An implementation of Matthew Honnibal's fast and accurate part-of-speech tagger based on the Averaged Perceptron
An implementation of Rob Pike's Virtual Machine-based regular expression engine