w13

1. 数据准备：收集数据与读取

​from sklearn.metrics import confusion_matrix
from sklearn.metrics import classification_report
from sklearn.naive_bayes import MultinomialNB
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction.text import TfidfVectorizer
import csv

 

2. 数据预处理：处理数据

file_path=r'SMSSpamCollectionjsn.txt'
a=open(file_path,'r',encoding='utf-8')
a_data=[]
a_label=[]
csv_reader=csv.reader(a,delimiter='\t')
for line in csv_reader:
    a_label.append(line[0])
    a_data.append(line[1])
a.close()

def preprocessing(text):
    preprocessing_text= text
    return preprocessing_text

3. 训练集与测试集：将先验数据按一定比例进行拆分。

X_train, X_test, Y_train, Y_test = train_test_split(a_data, a_label, test_size=0.3, random_state=0,stratify=a_label)
ectorizer=TfidfVectorizer(min_df=2,ngram_range=(1,2),stop_words='english',strip_accents='unicode',norm='l2')
X_train=vectorizer.fit_transform(X_train)
X_test=vectorizer.transform(X_test)
clf = MultinomialNB().fit(X_train,Y_train)
y_nb_pred = clf.predict(X_test)

4. 提取数据特征，将文本解析为词向量 。

from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test = train_test_split(sms_data,sms_label,test_size=0.3,random_state=0,stratify=sms_label)

from sklearn.feature_extraction.text import TfidfVectorizer
vectorizer=TfidfVectorizer(min_df=2,ngram_range=(1,2),stop_words='english',strip_accents='unicode',norm='12')
X_train=vectorizer.fit_transform(x_train)
X_test=vectorizer.transform(x_test)

5. 训练模型：建立模型，用训练数据训练模型。即根据训练样本集，计算词项出现的概率P(xi|y)，后得到各类下词汇出现概率的向量 。

from sklearn.naive_bayes import MultinomialNB
clf=MultinomialNB().fit(x_train,y_train)

6. 测试模型：用测试数据集评估模型预测的正确率。

混淆矩阵

准确率、精确率、召回率、F值

from sklearn.metrics import confusion_matrix
from sklearn.metrics import classification_report


cm=confusion_matrix(y_test.y_nb_pred)
print(cm)

cr=classification_report(y_test.y_nb_pred)
print(cr)

 

7. 预测一封新邮件的类别

new_email=['新邮件']

vectorizer(new_email)

clf.predict(new_email)