ku-nlp/jumanpp
Japanese morphological analyzer that uses RNNLM for semantic plausibility scoring of word sequences.
Not currently ranked — collecting fresh signals.
star history
Juman++ is a Japanese morphological analyzer and tokenizer that segments text and assigns part-of-speech tags. It employs a recurrent neural network language model (RNNLM) to evaluate semantic plausibility of word sequence interpretations, improving disambiguation accuracy over traditional rule-based analyzers. Version 2 offers substantially improved analysis speed while maintaining accuracy.
Frequently asked
- What is ku-nlp/jumanpp?
- Japanese morphological analyzer that uses RNNLM for semantic plausibility scoring of word sequences.
- Is jumanpp open source?
- Yes — ku-nlp/jumanpp is open source, released under the Apache-2.0 license.
- What language is jumanpp written in?
- ku-nlp/jumanpp is primarily written in C++.
- How popular is jumanpp?
- ku-nlp/jumanpp has 414 stars on GitHub.
- Where can I find jumanpp?
- ku-nlp/jumanpp is on GitHub at https://github.com/ku-nlp/jumanpp.