情感分析(支持向量機，邏輯迴歸)

以某寶購物評論爲例（表分爲好評和差評）

好評

df = pd.read_excel("F:/文本大數據/購物評論.xlsx",sheet_name="正向",header=None)
df

標記爲1

df['class']=1
df.head()

差評

df1 = pd.read_excel("F:/文本大數據/購物評論.xlsx",sheet_name="負向",header=None)
df1

標記爲0

df1['class']=0
df1.head()

合併

df2 = df.append(df1,ignore_index=True)
df2

分詞和預處理

import jieba
cuttxt = lambda x:" ".join(jieba.lcut(x))
df2["segment"]=df2[0].apply(cuttxt)
df2.head()

詞袋模型

from sklearn.feature_extraction.text import CountVectorizer

# 詞袋模型
countvec = CountVectorizer()
countvec = countvec.fit_transform(df2["segment"])
countvec

from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test = train_test_split(countvec,df2['class'],test_size=0)
x_train

from sklearn.svm import SVC
clf = SVC()# 支持向量機
clf.fit(x_train,y_train)

clf.score(x_train,y_train)

這裏會花較長的時間，需要耐心等待！

如果想提高準確率可以通過改變參數：

邏輯迴歸模型

from sklearn.linear_model import LogisticRegression
logistic = LogisticRegression()
logistic.fit(x_train,y_train)

logistic.score(x_train,y_train)

模型評估

SVM

from sklearn.metrics import classification_report
print(classification_report(y_test,clf.predict(x_test)))

s = """輸入你想預測的評論"""
s_seg = " ".join(jieba.lcut(s))
s_seg_vec = countvec.transform([s_seg])
result = clf.predict(s_seg_vec)
result

邏輯迴歸

from sklearn.metrics import classification_report
print(classification_report(y_test,logistic.predict(x_text)))

s = """輸入你想預測的評論"""
s_seg = " ".join(jieba.lcut(s))
s_seg_vec = countvec.transform([s_seg])
result = logistic.predict(s_seg_vec)
result