KoreanSentenceAnalyser

Build Status Hex pm

Docs: https://hexdocs.pm/korean_sentence_analyser.

How to use

Add to mix.exs

def deps do
  [
    {:korean_sentence_analyser, "~> 0.1.2"}
  ]
end

Start the python flask interface (mini api)

docker-compose up -d

For sentence analysis
KoreanSentenceAnalyser.analyse_sentence("한국어 배우기가 재미있어용")

Which will return a map, or nil if nothing is found

%{
    "tokens" => [
        %{"token" => "한국어", "type" => "Noun"},
        %{"token" => "배우다", "type" => "Verb"},
        %{"token" => "재미있다", "type" => "Adjective"}
    ]
}
For a single word analysis
KoreanSentenceAnalyser.get_the_stem_of_a_word("완벽했지")

which will return a string, or nil if empty

"완벽하다"

KoNLPy

Docs can be found at http://konlpy.org/en/latest/api

Debugging

Turn on debug=True in app.py

Note: Elasticsearch version

If you want to use the elasticsearch version, use the 1.0.0 tagged release on github, and 0.1.1 on Hex.pm