【机器学习】朴素贝叶斯文本分类案例

import pandas as pd
from sklearn.feature_extraction.text import CountVectorizer
import jieba
import numpy as np
from sklearn.naive_bayes import MultinomialNB

# 1、加载数据
data = pd.read_csv("./data.csv", encoding="ansi")
print("data:
", data)
print("data 的列索引：
", data.columns)

content = []

# 分词
for tmp in data.loc[:, "内容 "]:
    print(tmp)
    # 以精确模式对文章进行分词
    seg = jieba.cut(tmp, cut_all=False)
    # seg = jieba.cut(tmp, cut_all=True)
    seg_ = ",".join(seg)

    content.append(seg_)

# print(content)

data.loc[:, "内容 "] = content

print(data)

# 加载停止词
with open("./stopwords.txt", "r", encoding="utf-8") as f:
    stopwords = f.readlines()
    # 去除掉停止词前后的空白字符
    st_list = [tmp.strip() for tmp in stopwords]
# print(st_list)
# print(len(st_list))
# 停止词去重
st_list = list(set(st_list))
# print("去重之后的停止词长度：
", len(st_list))

# # 2、将文本内容转化为数值类型
# #  统计词数 统计词的重要性程度
# # 1、实例化对象
conv = CountVectorizer(stop_words=st_list)
# 2、词数统计
x = conv.fit_transform(data.loc[:, "内容 "])
print("x: 
", x)

# 获取统计的词语
feature_names = conv.get_feature_names()
res = x.toarray()
print(feature_names)
print("res:
", res)

# 将目标值获取到
data.loc[data.loc[:, "评价"] == "好评", "评价"] = 0
data.loc[data.loc[:, "评价"] == "差评", "评价"] = 1

# print(data.dtypes)
# 将目标值转化为Int类型
data.loc[:, "评价"] = data.loc[:, "评价"].astype(np.int64)
print("data的type: 
", data.dtypes)

# 将转化为数组之后的特征与目标值进行拼接
data = np.concatenate((res, data.loc[:, "评价"].values.reshape(-1, 1)), axis=1)

print("data:
", data)
print("data:
", data.dtype)

# 拆分成训练集与测试集
train = data[[0, 1, 3, 6, 8, 9, 10, 11, 12, 4], :]
test = data[[2, 5, 7], :]

# 构建朴素贝叶斯算法进行分类
# 1、实例化对象
nb = MultinomialNB(alpha=1.0)
# 2、训练数据
nb.fit(train[:, :-1], train[:, -1])
# 3、预测数据
y_predict = nb.predict(test[:, :-1])

# 获取准确率
score = nb.score(test[:, :-1], test[:, -1])

print("预测结果：
", y_predict)
print("准确率：
", score)