词条|笔者_Word2VecfastTextGlove训练词向量

篇首语：本文由编程笔记#小编为大家整理，主要介绍了Word2VecfastTextGlove训练词向量相关的知识，希望对你有一定的参考价值。

Word2Vec

环境&＃xff1a;
gensim 3.5.0
python 3.6.1

训练

import logging
import os.path
import sys
import multiprocessing
from gensim.corpora import WikiCorpus
from gensim.models import Word2Vec
from gensim.models.word2vec import LineSentence
program &＃61; os.path.basename(&＃39;train_word2vec_model&＃39;)
logger &＃61; logging.getLogger(program)
logging.basicConfig(format&＃61;&＃39;%(asctime)s: %(levelname)s: %(message)s&＃39;)
logging.root.setLevel(level&＃61;logging.INFO)
logger.info("running %s" % &＃39; &＃39;.join(sys.argv))
inp&＃61;&＃39;./data/cleaned_tweets_text_160W.csv&＃39;
outp&＃61;&＃39;Word2vec_model/word2vec_model_160w_200d&＃39;
n_dim&＃61;200
model &＃61; Word2Vec(LineSentence(inp), size&＃61;n_dim, window&＃61;5, min_count&＃61;5,
workers&＃61;multiprocessing.cpu_count())
model.save(outp)


使用

import gensim
outp&＃61;&＃39;Word2vec_model/word2vec_model_160w_200d&＃39;
model &＃61; gensim.models.Word2Vec.load(outp)


model[&＃39;happy&＃39;]


array([ -7.10430890e-02, -5.29273868e-01, 4.72357810e-01,
6.88659430e-01, -2.42118329e-01, -9.04537499e-01,
3.61087114e-01, -3.96869183e-01, -1.67573178e&＃43;00,
...
1.18314767e&＃43;00, -1.24723041e&＃43;00, 1.19374382e&＃43;00,
3.74429256e-01, 1.68333733e&＃43;00], dtype&＃61;float32)

result &＃61; model.most_similar(u"sad",topn&＃61;10)
for e in result:
print(e[0],e[1])


sad" 0.760986328125
upset 0.7119562029838562
sadd 0.7007699012756348
depressed 0.6737781763076782
bummed 0.6687890291213989
saad 0.6555933952331543
upsetting 0.6473105549812317
devastated 0.6356861591339111
disappointed 0.6215260624885559
heartbroken 0.5960309505462646

# man - woman &＃61; king - queen
# man - woman &＃61; husband - wife
# man - woman &＃61; boy - girl
model.most_similar([&＃39;woman&＃39;, &＃39;boy&＃39;], [&＃39;man&＃39;], topn&＃61;1)


[(‘girl’, 0.6189800500869751)]

print (model.doesnt_match(u"happy sad like desk".split()))


desk

fastText

环境&＃xff1a;
gensim 3.5.0
python 3.6.1

安装fasttext
可见&＃xff1a;https://github.com/facebookresearch/fastText/tree/master/python

$ git clone https://github.com/facebookresearch/fastText.git
$ cd fastText
$ pip install .


训练

from gensim.models import FastText
from gensim.models.word2vec import LineSentence
import logging
import os.path
import sys
program &＃61; os.path.basename(&＃39;fast-text&＃39;)
logger &＃61; logging.getLogger(program)
logging.basicConfig(format&＃61;&＃39;%(asctime)s: %(levelname)s: %(message)s&＃39;)
logging.root.setLevel(level&＃61;logging.INFO)
logger.info("running %s" % &＃39; &＃39;.join(sys.argv))
inp&＃61;&＃39;./data/cleaned_tweets_text_160W.csv&＃39;
outp&＃61;&＃39;fasttext_model/fasttext_model_160w_200d&＃39;
n_dim&＃61;200
model &＃61; FastText(LineSentence(inp), size&＃61;n_dim, window&＃61;5, min_count&＃61;5)
model.save(outp)


使用

from gensim.models import FastText
# 使用&＃xff1a;模型得保存与加载
outp&＃61;&＃39;../word-Vectorization/fasttext_model/fasttext_model_160w_200d&＃39;
model &＃61; FastText.load(outp)


# 指定词条词向量
model[&＃39;happy&＃39;]


array([ 8.10116410e-01, 5.25403082e-01, -3.36705118e-01,
-4.42356855e-01, -2.04296446e&＃43;00, -9.93114054e-01,
…
1.20443761e&＃43;00, -3.73263896e-01, 1.15821147e&＃43;00,
-1.95465660e&＃43;00, 2.11104417e&＃43;00], dtype&＃61;float32)

result &＃61; model.most_similar(u"queeen")
for e in result:
print(e[0],e[1])


queen" 0.8662827014923096
mcqueen 0.8575066924095154
queenie 0.8428698182106018
queens 0.7900538444519043
queer 0.7410856485366821
queezy 0.7156778573989868
queenstown 0.7068071365356445
quen 0.7057324647903442
queensland 0.7055833339691162

Glove

环境&＃xff1a;

python 3.6.1

官方glove:https://github.com/stanfordnlp/GloVe
笔者使用&＃xff1a;https://github.com/maciejkula/glove-python

安装&＃xff1a;
pip install glove_python

训练

from __future__ import print_function
import argparse
import pprint
import gensim
from glove import Glove
from glove import Corpus
from gensim.models.word2vec import LineSentence
inp&＃61;&＃39;./data/processed_160Msemeval-2016-2017-task3-QAText&＃39;
outp&＃61;&＃39;glove_model/QAText_200d&＃39;
n_dim&＃61;200
corpus_model &＃61; Corpus()
corpus_model.fit(LineSentence(inp), window&＃61;5)
#corpus_model.save(&＃39;corpus.model&＃39;)
print(&＃39;Dict size: %s&＃39; % len(corpus_model.dictionary))
print(&＃39;Collocations: %s&＃39; % corpus_model.matrix.nnz)
glove &＃61; Glove(no_components&＃61;n_dim, learning_rate&＃61;0.05)
glove.fit(corpus_model.matrix, epochs&＃61;10,
no_threads&＃61;1, verbose&＃61;True)
glove.add_dictionary(corpus_model.dictionary)
glove.save(outp)


使用

from glove import Glove
outp&＃61;&＃39;../word-Vectorization/glove_model/glove_model_160w_200d&＃39;
glove &＃61; Glove.load(outp)


glove.most_similar(&＃39;happiness&＃39;, number&＃61;10)


[(‘optimism’, 0.88474224538009472),
(‘technology’, 0.87399374238355365),
(‘photography’, 0.87293168487727013),
(‘jquery’, 0.87160586331482937),
(‘excitement’, 0.87015550639464156),
(‘mailplane’, 0.86709197956363315),
(‘secrets’, 0.86616698912523094),
(‘laughter’, 0.86593359775060097),
(‘records’, 0.86521431383969472)]

# 指定词条词向量
glove.word_vectors[glove.dictionary[&＃39;happy&＃39;]]


array([ 0.1824374 , -0.15493986, -0.23131742, -0.20251903, -0.25899053,
0.16043589, -0.11017494, -0.15413852, 0.12485044, 0.28871841,
…
0.14323183, -0.21197602, -0.18841062, 0.32888953, -0.32943953,
0.15334943, -0.09995708, 0.10678763, 0.12507708, -0.26995188,
0.17373759, -0.17477675, 0.16042781, -0.3823496 , -0.21795925])

参考&＃xff1a;
https://www.zybuluo.com/hanxiaoyang/note/472184?tdsourcetag&＃61;s_pctim_aiomsg
https://blog.csdn.net/sinat_26917383/article/details/83041424
https://blog.csdn.net/sinat_26917383/article/details/83029140