文本处理

 

#导入os包加载数据目录
import os
path = r'E:\dzy'
#停词库
with open(r'e:\\stopsCN.txt', encoding='utf-8') as f:
    stopwords = f.read().split('\n')

　　

#对数据进行标准编码处理(encoding='utf-8')
import codecs
import jieba
#存放文件名
filePaths = []
#存放读取的数据
fileContents = []
#存放文件类型
fileClasses = []
#进行遍历实现转码读取处理并对每条新闻进行切分
for root, dirs, files in os.walk(path): 
    for name in files:
        filePath = os.path.join(root, name)
        filePaths.append(filePath)
        fileClasses.append(filePath.split('\\')[2])
        f = codecs.open(filePath, 'r', 'utf-8') 
        fileContent = f.read()
        fileContent = fileContent.replace('\n','')
        tokens = [token for token in jieba.cut(fileContent)]
        tokens = " ".join([token for token in tokens if token not in stopwords])
        f.close()
        fileContents.append(tokens)

　　

import pandas;
all_datas = pandas.DataFrame({
    'fileClass': fileClasses, 
    'fileContent': fileContents
})
print(all_datas)

　　

 

str=''
for i in range(len(fileContents)):
    str+=fileContents[i]
#TF-IDF算法
#统计词频
import jieba.analyse
keywords = jieba.analyse.extract_tags(str, topK=20, withWeight=True, allowPOS=('n','nr','ns'))
print(keywords )

　　

from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test = train_test_split(fileContents,fileClasses,test_size=0.3,random_state=0,stratify=fileClasses)
vectorizer = TfidfVectorizer()
X_train = vectorizer.fit_transform(x_train)
X_test = vectorizer.transform(x_test)
from sklearn.naive_bayes import  MultinomialNB
clf= MultinomialNB().fit(X_train,y_train)
y_nb_pred=clf.predict(X_test)
#分类结果显示
from sklearn.metrics import confusion_matrix
from sklearn.metrics import classification_report

print('nb_confusion_matrix:')
print(y_nb_pred.shape,y_nb_pred)#x_test预测结果
cm=confusion_matrix(y_test,y_nb_pred)#混淆矩阵
print('nb_classification_report:')
print(cm)