Metadata-Version: 2.0
Name: soynlp
Version: 0.0.24
Summary: Unsupervised Korean Natural Language Processing Toolkits
Home-page: https://github.com/lovit/soynlp
Author: Lovit
Author-email: soy.lovit@gmail.com
License: UNKNOWN
Keywords: korean natural language processing
Platform: UNKNOWN
Requires-Dist: numpy (>=1.12.1)
Requires-Dist: psutil (>=5.0.1)

It contains unsupervised word extraction, tokenizers and noun extractors. 
These algorithms are not depending training corpus but extract patterns from data by theirselves.

Current version has follows
- Word extraction
  - Cohesion score
  - Branching Entropy
  - Accessor Variety
- Tokenizers
  - RegexTokenizer
  - LTokenizer
  - MaxScoreTokenizer
- Noun extractor
  - LRNounExtractor


Following packages are helpful
- krwordrank: Unsupervised Korean word/keyword extractor
  - https://github.com/lovit/KR-WordRank
  - pip install krwordrank
- soyspacing: Korean spacing error corrector
  - https://github.com/lovit/soyspacing
  - pip install soyspacing


