gensim---gensim构建词向量模型Word2Vec

利用库gensim构建向量模型Word2Vec

数据准备格式

语料：每句话内容，词之间相互以空格隔开
模型构建代码：

from gensim.models import word2vec

class Solution():
    def __init__(self):
        # 语料路径
        self.corpus_path = r"xxx\corpus.txt"
        # 模型保存路径
        self.save_path = r"xxx\word2vec_sg.model"
        # 模型加载路径
        self.model_path = r"xxx\word2vec_sg.model"
        
    def buildWord2Vec(self):
        # 获取训练语料
        print('new_energy_sentences:')
        sentences = word2vec.Text8Corpus(self.corpus_path)
    
        # 模型训练
        print('Model train...')
        model = word2vec.Word2Vec(sentences,sg=1,min_count=3, size=100, window=8)
    
        print('Model save...')
        model.save(self.save_path)
        
    def testModel(self):
        model = word2vec.Word2Vec.load(self.model_path)
        # 计算两个词的相似度/相关程度
        print(model.similarity("喜剧", "惊悚"))
        
        # 计算某个词的相关词列表
        print(model.most_similar("动作", topn=10))  # 10个最相关的
        
    def main(self):
        self.buildWord2Vec()
        # self.testModel()
        pass
ob = Solution()
ob.main()