过滤垃圾邮件贝叶斯python代码

贝叶斯算法是一种常用的过滤垃圾邮件的方法，以下是一个简单的用Python实现的贝叶斯垃圾邮件过滤代码示例：

import re
import os
import math

spam_words = {}
ham_words = {}

def tokenize(text):
    words = re.findall(r'\b\w+\b', text.lower())
    return set(words)

def train_spam(directory):
    for filename in os.listdir(directory):
        with open(os.path.join(directory, filename), 'r', encoding='utf-8') as file:
            text = file.read()
            words = tokenize(text)
            for word in words:
                spam_words[word] = spam_words.get(word, 0) + 1

def train_ham(directory):
    for filename in os.listdir(directory):
        with open(os.path.join(directory, filename), 'r', encoding='utf-8') as file:
            text = file.read()
            words = tokenize(text)
            for word in words:
                ham_words[word] = ham_words.get(word, 0) + 1

def calculate_spam_prob(text):
    words = tokenize(text)
    spam_prob = 0
    for word in words:
        spam_count = spam_words.get(word, 0)
        ham_count = ham_words.get(word, 0)
        if spam_count + ham_count > 0:
            spam_prob += math.log((spam_count + 0.5) / (spam_count + ham_count + 1))
    return spam_prob

def classify_email(text):
    spam_prob = calculate_spam_prob(text)
    if spam_prob > 0:
        return "垃圾邮件"
    else:
        return "正常邮件"

# 训练垃圾邮件
train_spam("spam_directory")
# 训练正常邮件
train_ham("ham_directory")

# 邮件分类
email_text = "这是一封垃圾邮件"
classification = classify_email(email_text)
print(classification)

以上代码中，`tokenize`函数用于将文本分词，`train_spam`和`train_ham`函数分别用于训练垃圾邮件和正常邮件，`calculate_spam_prob`函数用于计算邮件的垃圾邮件概率，`classify_email`函数用于判断邮件是否为垃圾邮件。使用时，需要将实际的垃圾邮件和正常邮件分别放置在`spam_directory`和`ham_directory`文件夹中，并将待分类的邮件文本传入`classify_email`函数进行分类。