基于Python 朴素贝叶斯--文本分类
# coding: utf-8
利用jupter book在线运行code。
步骤:
准备分类文档内容和分类标签,停用词文档
利用Jieba(中文)/NTLK(英文)将文档中单词分词
加载停用词文件,生成TFIDF向量,计算单词的TFIDF,(TF:词频,IDF:逆向文档频率=
(文档数/(单词出现的文档数+1))
使用多项式贝叶斯算法生成分类器
预测结果并计算分类器的准确率
# 中文文本分类
import os
import jieba
import warnings
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.naive_bayes import MultinomialNB
from sklearn import metrics
def cut_words(file_path):
"""
对文本进行切词
:param file_path: txt文本路径
:return: 用空格分词的字符串
"""
text_with_spaces = ''
text=open(file_path, 'r', encoding='gb18030').read()
textcut = jieba.cut(text)
for word in textcut:
text_with_spaces += word +' '
return text_with_spaces
def loadfile(file_dir, label):
"""
将路径下的所有文件加载
:param file_dir: 保存txt文件目录
:param label: 文档标签
:return: 分词后的文档列表和标签
"""
file_list = os.listdir(file_dir)
words_list = []
labels_list = []
for file in file_list:
file_path = file_dir + '/' + file
words_list.append(cut_words(file_path))
labels_list.append(label)
return words_list, labels_list
pathdir=...
# 训练数据
train_words_list1, train_labels1 = loadfile(pathdir+'text classification/train/女性', '女性')
train_words_list2, train_labels2 = loadfile(pathdir+'text classification/train/体育', '体育')
train_words_list3, train_labels3 = loadfile(pathdir+'text classification/train/文学', '文学')
train_words_list4, train_labels4 = loadfile(pathdir+'text classification/train/校园', '校园')
train_words_list = train_words_list1 + train_words_list2 + train_words_list3 + train_words_list4
train_labels = train_labels1 + train_labels2 + train_labels3 + train_labels4
# 测试数据
test_words_list1, test_labels1 = loadfile(pathdir+'text classification/test/女性', '女性')
test_words_list2, test_labels2 = loadfile(pathdir+'text classification/test/体育', '体育')
test_words_list3, test_labels3 = loadfile(pathdir+'text classification/test/文学', '文学')
test_words_list4, test_labels4 = loadfile(pathdir+'text classification/test/校园', '校园')
test_words_list = test_words_list1 + test_words_list2 + test_words_list3 + test_words_list4
test_labels = test_labels1 + test_labels2 + test_labels3 + test_labels4
#加载停用词
stop_words = open(pathdir+'text classification/stop/stopword.txt', 'r', encoding='utf-8').read()
print(stop_words)
stop_words = stop_words.encode('utf-8').decode('utf-8-sig') # 列表头部\ufeff处理
stop_words = stop_words.split('\n') # 根据分隔符分隔
new_stopword=['ain','aren', 'couldn', 'didn', 'doesn', 'don', 'hadn', 'hasn', 'haven', 'isn', 'll', 'mon', 'shouldn', 've', 'wasn', 'weren', 'won', 'wouldn']
stop_words=stop_words.append(new_stopword)
print(stop_words)
# 计算单词权重
tf2 = TfidfVectorizer(stop_words=stop_words, max_df=0.5)
train_features = tf2.fit_transform(train_words_list)
print(train_features)
print('不重复的词:',len(tf2.get_feature_names()))
print('每个单词的ID:', tf2.vocabulary_)
print('每个单词的tfidf值:', train_features.toarray())
# 上面fit过了,这里transform
test_features = tf2.transform(test_words_list)
# 多项式贝叶斯分类器
from sklearn.naive_bayes import MultinomialNB
clf = MultinomialNB(alpha=0.001).fit(train_features, train_labels)
predicted_labels=clf.predict(test_features)
# 计算准确率
print('准确率为:', metrics.accuracy_score(test_labels, predicted_labels))
#准确率为:0.92