実験結果: https://docs.google.com/spreadsheets/d/1l_OqjuCC2D44KvFFDYHDeN7GRReRS9nh7JXytctfxHw/edit#gid=0
cython implementation of word2vec
$ git clone https://github.com/jyori112/cython-word2vec
$ cd cython-word2vec
$ ./cythonize.sh; python setup.py develop
Create dictionary first,
$ python main.py build_dic [corpus] [dictionary]
Train word embeddings
$ python main.py train [corpus] [dictionary] [output]