1. 读邮件数据集文件，提取邮件本身与标签。

列表

numpy数组

2.邮件预处理

  1. 邮件分句
  2. 句子分词
  3. 大小写，标点符号，去掉过短的单词
  4. 词性还原：复数、时态、比较级
  5. 连接成字符串

2.1 传统方法来实现

2.2 nltk库的安装与使用

pip install nltk

import nltk

nltk.download()     # sever地址改成 http://www.nltk.org/nltk_data/

或

https://github.com/nltk/nltk_data下载gh-pages分支，里面的Packages就是我们要的资源。

将Packages文件夹改名为nltk_data。

或

网盘链接：https://pan.baidu.com/s/1iJGCrz4fW3uYpuquB5jbew    提取码：o5ea

放在用户目录。

----------------------------------

安装完成，通过下述命令可查看nltk版本：

import nltk

print nltk.__doc__

 

 

 

 

 

2.1 nltk库 分词

nltk.sent_tokenize(text) #对文本按照句子进行分割

nltk.word_tokenize(sent) #对句子进行分词

2.2 punkt 停用词

from nltk.corpus import stopwords

stops=stopwords.words('english')

*如果提示需要下载punkt

nltk.download(‘punkt’)

或 下载punkt.zip

https://pan.baidu.com/s/1OwLB0O8fBWkdLx8VJ-9uNQ  密码：mema

复制到对应的失败的目录C:\Users\Administrator\AppData\Roaming\nltk_data\tokenizers并解压。

 

2.3 NLTK 词性标注

nltk.pos_tag(tokens)

2.4 Lemmatisation（词性还原）

from nltk.stem import WordNetLemmatizer

lemmatizer = WordNetLemmatizer()

lemmatizer.lemmatize('leaves') #缺省名词

lemmatizer.lemmatize('best',pos='a')

lemmatizer.lemmatize('made',pos='v')

一般先要分词、词性标注，再按词性做词性还原。

2.5 编写预处理函数

def preprocessing(text):

sms_data.append(preprocessing(line[1])) #对每封邮件做预处理

3. 训练集与测试集

4. 词向量

5. 模型

代码：

import numpy as np

def loadDataSet():
   postingList = [
          ['my','dog','has','flea','problems','help','please'],
          ['maybe','not','take','him','to','dog','park','stupid'],
          ['my','dalmation','is','so','cute','I','love','him'],
          ['stop','posting','stupid','worthless','garbage'],
          ['mr','licks','ate','my','steak','how','to','stop','him'],
          ['quit','buying','worthless','dog','food','stupid']
                 ]

   classVec = [0,1,0,1,0,1]  #1:侮辱性文字，0:正常言论

   return postingList,classVec

def createVocabList(dataSet):
   vocabSet = set([])        #创建一个空集
   for document in dataSet:
      vocabSet = vocabSet | set(document)  #创建两个集合的并集，去重操作
   return list(vocabSet)

def setOfWords2Vec(vocabList,inputSet):    #对单词做简单的词向量
   returnVec = [0] * len(vocabList)
   for word in inputSet:
      if word in vocabList:
         returnVec[vocabList.index(word)] = 1
      else:
         print ("the word: %s is not in my Vocabulary!" % word)

   return returnVec

#训练函数
def trainNB0(trainMatrix,trainCategory):
   #trainMatrix     : 0,1表示的文档矩阵
   #trainCategory   : 类别标签构成的向量

   numTrainDocs = len(trainMatrix)
   numWords     = len(trainMatrix[0])
   pAbusive     = sum(trainCategory) / float(numTrainDocs)  #P(c1)
   p0Num        = np.zeros(numWords)
   p1Num        = np.zeros(numWords)
   p0Denom      = 0.0
   p1Denom      = 0.0
   #len(trainMatrix)  == len(trainCategory)
   for i in range(numTrainDocs):
      if trainCategory[i] == 1:
         p1Num    += trainMatrix[i]
         p1Denom  += sum(trainMatrix[i])
      else:
         p0Num     += trainMatrix[i]
         p0Denom   += sum(trainMatrix[i])

   p0Vect = p0Num / p0Denom                      #P(Wi|C1)的向量形式
   p1Vect = p1Num / p1Denom                      #p(wi|c2)的向量形式
   return p0Vect,p1Vect,pAbusive

listPosts,listClasses = loadDataSet()
myVocabList = createVocabList(listPosts)

trainMat = []
for postinDoc in listPosts:
   trainMat.append(setOfWords2Vec(myVocabList,postinDoc))

print (trainMat)

p0V,p1V,pAb = trainNB0(trainMat,listClasses)
print ("p0V: \n",p0V)
print ("p1V: \n",p1V)
print ("pAb: \n",pAb)


 运行结果：