gensim_word_vector.md

Gensim Word2Vec - word vector

• 불러와서 사용하는 기본 예제 코드

  from gensim.models import Word2Vec
  model = Word2Vec.load("word_embedding.model")
  
  len(model.wv.vocab)
  
  model.wv.similarity('회사/NNG', '기업/NNG')
  model.wv.similarity('국립/NNG', '사립/NNG')
  print(model.wv.most_similar("국립/NNG", topn=10))
  print(model.wv.most_similar(positive=["젊은/NNG", "사람/NNG"], topn=10))
  print(model.wv.most_similar(positive=["젊은/NNG", "사람/NNG"], negative=["한국/NNG"], topn=10))
  
  vector = model.wv['국립/NNG']
  print(vector)

• FutureWarning: Conversion of the second argument of issubdtype from int to np.signedinteger is deprecated. In future, it will be treated as np.int64 == np.dtype(int).type. if np.issubdtype(vec.dtype, np.int):

  • pip install numpy==1.13.0
  • pip install h5py==2.8.0rc1
  • pip install --upgrade ipykernel
  • 별거 다 해봤지만 나한텐 안됨ㅠ 그래서 결국 warning을 지웠다.

  import warnings
  warnings.filterwarnings('ignore')