樸素貝葉斯應用：垃圾郵件分類

代碼：

import csv
# 讀數據
file_path = r‘EmailData.txt‘
EmailData = open(file_path,‘r‘,encoding=‘utf-8‘)
Email_data = []
Email_target = []
csv_reader = csv.reader(EmailData,delimiter=‘\t‘)
# 將數據分別存入數據列表和目標分類列表
for line in csv_reader:
    Email_data.append(line[1])
    Email_target.append(line[0])
EmailData.close()

 
# 把無意義的符號都替換成空格
Email_data_clear = []
for line in Email_data:
    # line :‘Go until jurong point, crazy.. Available only in bugis n great world la e buffet...‘
    # 每一行都去掉無意義符號並按空格分詞
    for char in line:
        if char.isalpha() is False:
            # 不是字母，發生替換操作:
            newString = line.replace(char,"
 
 ")
    tempList = newString.split(" ")
    # 將處理好後的一行數據追加到存放幹凈數據的列表
    Email_data_clear.append(tempList)

# 去掉長度不大於3的詞和沒有語義的詞
Email_data_clear2 = []
for line in Email_data_clear:
    tempList = []
    for word in line:
        if word != ‘‘ and len(word) > 3 and word.isalpha():
            tempList.append(word)
    tempString 
 
= ‘ ‘.join(tempList)
    Email_data_clear2.append(tempString)
Email_data_clear = Email_data_clear2

# 將數據分為訓練集和測試集
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test = train_test_split(Email_data_clear2,Email_target,test_size=0.3,random_state=0,stratify=Email_target)

# 建立數據的特征向量
from sklearn.feature_extraction.text import TfidfVectorizer
tfidf = TfidfVectorizer()
X_train = tfidf.fit_transform(x_train)
X_test = tfidf.transform(x_test)

# 觀察向量
import numpy as np
X_train = X_train.toarray()
X_test = X_test.toarray()
X_train.shape
# 輸出不為0的列
for i in range(X_train.shape[0]):
    for j in range(X_train.shape[1]):
        if X_train[i][j] != 0:
            print(i,j,X_train[i][j])

# 建立模型
from sklearn.naive_bayes import GaussianNB
gnb = GaussianNB()
module = gnb.fit(X_train,y_train)
y_predict = module.predict(X_test)

# 輸出模型分類的各個指標
from sklearn.metrics import classification_report
cr = classification_report(y_predict,y_test)
print(cr)

截圖：

清洗後的數據：

特征向量：

模型指標：

第十二次作業