[NLP]LDA主题模型的python实现

# -*- coding:utf-8 -*-
"
Created on Mon Aug 19 14:56:19 2019
@author: Luxuriant
"

import numpy as np
from gensim import corpora, models, similarities
from pprint import pprint
import time

# import logging
# logging.basicConfig(format=‘%(asctime)s : %(levelname)s : %(message)s‘, level=logging.INFO)

def load_stopword():
    ‘‘‘
    加载停用词表
    :return: 返回停用词的列表
    ‘‘‘
    f_stop = open(‘stopwords.txt‘, encoding=‘utf-8‘)
    sw = [line.strip() for line in f_stop]
    f_stop.close()
    return sw

if __name__ == ‘__main__‘:

    print(‘1.初始化停止词列表 ------‘)
    # 开始的时间
    t_start = time.time()
    # 加载停用词表
    stop_words = load_stopword()

    print(‘2.开始读入语料数据 ------ ‘)
    # 读入语料库
    f = open(‘乡贤形象文本-cutfile.txt‘, encoding=‘utf-8‘)
    # 语料库分词并去停用词
    texts = [[word for word in line.strip().lower().split() if word not in stop_words] for line in f]

    print(‘读入语料数据完成，用时%.3f秒‘ % (time.time() - t_start))
    f.close()
    M = len(texts)
    print(‘文本数目：%d个‘ % M)

    print(‘3.正在建立词典 ------‘)
    # 建立字典
    dictionary = corpora.Dictionary(texts)
    V = len(dictionary)

    print(‘4.正在计算文本向量 ------‘)
    # 转换文本数据为索引，并计数
    corpus = [dictionary.doc2bow(text) for text in texts]

    print(‘5.正在计算文档TF-IDF ------‘)
    t_start = time.time()
    # 计算tf-idf值
    corpus_tfidf = models.TfidfModel(corpus)[corpus]
    print(‘建立文档TF-IDF完成，用时%.3f秒‘ % (time.time() - t_start))

    print(‘6.LDA模型拟合推断 ------‘)
    # 训练模型
    num_topics = 30
    t_start = time.time()
    lda = models.LdaModel(corpus_tfidf, num_topics=num_topics, id2word=dictionary,
                      alpha=0.01, eta=0.01, minimum_probability=0.001,
                      update_every=1, chunksize=100, passes=1)
    print(‘LDA模型完成，训练时间为\t%.3f秒‘ % (time.time() - t_start))

    # 随机打印某10个文档的主题
    num_show_topic = 10  # 每个文档显示前几个主题
    print(‘7.结果：10个文档的主题分布：--‘)
    doc_topics = lda.get_document_topics(corpus_tfidf)  # 所有文档的主题分布
    idx = np.arange(M)
    np.random.shuffle(idx)
    idx = idx[:10]
    for i in idx:
        topic = np.array(doc_topics[i])
        topic_distribute = np.array(topic[:, 1])
        # print topic_distribute
        topic_idx = topic_distribute.argsort()[:-num_show_topic - 1:-1]
        print(‘第%d个文档的前%d个主题：‘ % (i, num_show_topic)), topic_idx
        print(topic_distribute[topic_idx])

    num_show_term = 10  # 每个主题显示几个词
    print(‘8.结果：每个主题的词分布：--‘)
    for topic_id in range(num_topics):
        print(‘主题#%d：\t‘ % topic_id)
        term_distribute_all = lda.get_topic_terms(topicid=topic_id)
        term_distribute = term_distribute_all[:num_show_term]
        term_distribute = np.array(term_distribute)
        term_id = term_distribute[:, 0].astype(np.int)
        print(‘词：\t‘, )
        for t in term_id:
            print(dictionary.id2token[t], )
        print(‘\n概率：\t‘, term_distribute[:, 1])