academiasinicanlplab / emotion_push

EmotionPush

Environment Setting

git submodule update --init corpus word2vec
cd word2vec
# If you are lwku lab member:
ln -s /corpus/wordvector/word2vec/google_word2vec_pretrained google_word2vec_pretrained
ln -s /corpus/wordvector/word2vec/google_word2vec_pretrained.syn0.npy google_word2vec_pretrained.syn0.npy
# Else, follow word2vec/README.md to setup word2vec

Source Tree Architecture

word2vec: word vector modules, import this modules to load pre-trained word vectors. corpus: raw corpus data and preprocessing script, read corpus/README.md for modre information. cache: cached data, e.g. word vector for word of certain corpus (loading word vectors for all words every time is not a good idea). model: trained model files. Running emotion push server requires at least one trained model. src: source code, read src/README.md for more information.