【一起从0开始学习人工智能0x02】字典特征抽取、文本特征抽取、中文文本特征抽取

sklearn.featurn_extraction()

sklearn.featurn_extraction.DicVertorizer(sparse=True,...)

#1）实例化一个转换器类 2)调用fir_transform()方法
from sklearn.feature_extraction import  DictVectorizer#导包
#下面的data是数据
data=[{'city':'北京','tempreature':100},
      {'city':'上海','tempreature':60},
      {'city':'深圳','tempreature':30},]
#1实例化一个转换器类
transfer=DictVectorizer()
#2调用一fit_transform()方法
data_new=transfer.fit_transform(data)
print("看一下转换的结果data_new:\n",data_new)
print("看一下特征的名字:\n",transfer.get_feature_names())

(0, 1) 1.0
(0, 3) 100.0
(1, 0) 1.0
(1, 3) 60.0
(2, 2) 1.0
(2, 3) 30.0

data_new 
 [[  0.   1.   0. 100.]
 [  1.   0.   0.  60.]
 [  0.   0.   1.  30.]]

from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction import DictVectorizer
from sklearn.feature_extraction.text import CountVectorizer
def datasets_demo():
    #获取数据集
    iris = load_iris();
    print('鸢尾花数据集的描述：\n', iris.data, iris.data.shape)
    x_train, x_test, y_train, y_test=train_test_split(iris.data, iris.target, test_size=0.2, random_state=22)
    print("xunlianjitezhengzhi:\n:", x_train, x_train.shape)
    return None
def count_demo():
    data = ["The voice of one man is the voice of no one one",
            "The voice of one woman is the voice of  one"]
    # 1.实例化一个转换器类
    transfer = CountVectorizer()  # 注意CountVecotrizer()没有sparse参数，所以无法通过调节sparse参数来调节返回的结果
    # transfer = CountVectorizer(stop_words=['one']) # 注意CountVecotrizer()有一个stop_words参数用于去除不需要提取的词，且stop_words必须传一个列表对象
    # 2.调用实例对象的fit_transform()
    new_data = transfer.fit_transform(data)  # 会自动去除单个词
    # print("new_data:\n",new_data) # 返回稀疏矩阵（通常默认情况下都是返回稀疏矩阵）
    print("new_data:\n", new_data.toarray())  # 返回密集矩阵
    return None
if __name__ == "__main__":
    # datasets_demo()
    count_demo()

new_data:
 [[1 1 1 2 3 2 2 0]
 [1 0 0 2 2 2 2 1]]
Process finished with exit code 0

import sklearn.feature_extraction.text as text
import jieba
transfer = text.CountVectorizer(stop_words=['vb'])
def count_chinese_demo2():
    data = ['发表回复这件事', '飞机里面飞一杯飞机专属奶茶', '没有什么比在飞机上喝一杯飞机专属的飞机奶茶要更好了']
    data_new = []
    # 中文文本分词
    for send in data:
        data_new.append(' '.join(list(jieba.cut(send))))
    print(data_new)
    # 文本特征提取
    data_final = transfer.fit_transform(data_new)
    print(data_final.toarray())
   # print(transfer.get_feature_names())
if __name__ == "__main__":
    count_chinese_demo2()

['发表 回复 这件 事', '飞机 里面 飞 一杯 飞机 专属 奶茶', '没有 什么 比 在 飞机 上 喝一杯 飞机 专属 的 飞机 奶茶 要 更好 了']
[[0 0 0 1 0 1 0 0 0 1 0 0]
 [1 1 0 0 0 0 1 0 0 0 1 2]
 [0 1 1 0 1 0 1 1 1 0 0 3]]
Process finished with exit code 0