贝叶斯算法实例源代码.pdfVIP

贝叶斯算法实例源代码

以下是一个基于贝叶斯算法的垃圾邮件分类的实例源代码：

python

importos

importstring

fromcollectionsimportCounter

classSpamFilter:

def__init__(self,spam_dir,ham_dir):

self.spam_dir=spam_dir

self.ham_dir=ham_dir

self.spam_words=Counter()

self.ham_words=Counter()

self.spam_total_words=0

self.ham_total_words=0

self.p_spam=0

self.p_ham=0

self.train()

deftrain(self):

#训练

forpathinos.listdir(self.spam_dir):

withopen(os.path.join(self.spam_dir,path),r,

encoding=ISO-8859-1)asf:

forlineinf.readlines():

string.punctuation)).split()

words=[w.lower()forwinwords]

self.spam_words.update(words)

self.spam_total_words+=len(words)

forpathinos.listdir(self.ham_dir):

withopen(os.path.join(self.ham_dir,path),r,

encoding=ISO-8859-1)asf:

forlineinf.readlines():

string.punctuation)).split()

words=[w.lower()forwinwords]

self.ham_words.update(words)

self.ham_total_words+=len(words)

self.p_spam=len(os.listdir(self.spam_dir))/

(len(os.listdir(self.spam_dir))+len(os.listdir(self.ham_dir)))

self.p_ham=1-self.p_spam

defclassify(self,text):

#分类

string.punctuation)).split()

words=[w.lower()forwinwords]

p_word_spam=1

p_word_ham=1

forwinwords:

p_word_spam*=(self.spam_words[w]+1)/

(self.spam_total_words+len(self.spam_words))

p_word_ham*=(self.ham_words[w]+1)/

(self.ham_total_words+len(self.ham_words))