Gensim的model使用word2vec 示例

# coding=utf-8
"""
 @ File: word2vec_gensim.py
 @Software: PyCharm
 @desc:
"""
from gensim.models import word2vec
import logging
logging.basicConfig(format='%(asctime)s: %(levelname)s: %(message)s', level=logging.INFO)
raw_sentences = ['the quick brown fox jumps over ther lazy dogs', 'yoyoyo you go home now to sleep']

sentences = [s.split() for s in raw_sentences]
print(sentences)
# out: [['the', 'quick', 'brown', 'fox', 'jumps', 'over', 'ther', 'lazy', 'dogs'], ['yoyoyo', 'you', 'go', 'home', 'now', 'to', 'sleep']]

# 传参是文章分词后的列表，每篇文章一个元素
model = word2vec.Word2Vec(sentences, min_count=1)

model.wv.save('m2.mdl')
# 或者
model.save('m1.mdl')

# 加载使用模型
md = word2vec.Word2Vec.load('m1.mdl')
# 用于比较单个词语
print(md.similarity('dogs', 'you'))
# out: -0.06432766
# wv是4.0新版本后的方法，代替model.n_similartity
# n_similarity用于比较文章
print(md.wv.n_similarity(['fox','dogs'], ['dogs', 'fox']))
# out：1.0

# most_similar找到相似度最高的词
word = 'dogs'
# 如果 word在词向量词库中
if word in model.wv.index2word:
 print(model.most_similar('dogs'))
else:
 print(word + ' is not in model')
 # 打印出词向量库中的所有词
 print(model.wv.index2word)

if model.__contains__(word):
 print(word + " is in model")

ref : https://blog.csdn.net/luoluonuoyasuolong/article/details/107810578