对于一个简单的文本情感分类来说，其实就是一个二分类，这篇博客主要讲述的是使用scikit-learn来做文本情感分类。分类主要分为两步：1)训练，主要根据训练集来学习分类模型的规则。2)分类，先用已知的测试集评估分类的准确率等，如果效果还可以，那么该模型对无标注的待测样本进行预测。
首先先介绍下我样本集，样本是已经分好词的酒店评论，第一列为标签，第二列为评论，前半部分为积极评论，后半部分为消极评论，格式如下：
下面实现了SVM,NB,逻辑回归，决策树，逻辑森林，KNN 等几种分类方法，主要代码如下：
#coding:utf-8
from matplotlib import pyplot
import scipy as sp
import numpy as np
from sklearn.cross_validation import train_test_split
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics import precision_recall_curve
from sklearn.metrics import classification_report
from numpy import *
#========SVM========#
def SvmClass(x_train, y_train):
from sklearn.svm import SVC
#调分类器
clf = SVC(kernel = 'linear',probability=True)#default with 'rbf'
clf.fit(x_train, y_train)#训练，对于监督模型来说是 fit(X, y)，对于非监督模型是 fit(X)
return clf
#=====NB=========#
def NbClass(x_train, y_train):
from sklearn.naive_bayes import MultinomialNB
clf=MultinomialNB(alpha=0.01).fit(x_train, y_train)
return clf
#========Logistic Regression========#
def LogisticClass(x_train, y_train):
from sklearn.linear_model import LogisticRegression
clf = LogisticRegression(penalty='l2')
clf.fit(x_train, y_train)
return clf
#========KNN========#
def KnnClass(x_train,y_train):
from sklearn.neighbors import KNeighborsClassifier
clf=KNeighborsClassifier()
clf.fit(x_train,y_train)
return clf
#========Decision Tree ========#
def DccisionClass(x_train,y_train):
from sklearn import tree
clf=tree.DecisionTreeClassifier()
clf.fit(x_train,y_train)
return clf
#========Random Forest Classifier ========#
def random_forest_class(x_train,y_train):
from sklearn.ensemble import RandomForestClassifier
clf= RandomForestClassifier(n_estimators=8)#参数n_estimators设置弱分类器的数量
clf.fit(x_train,y_train)
return clf
#========准确率召回率 ========#
def Precision(clf):
doc_class_predicted = clf.predict(x_test)
print(np.mean(doc_class_predicted == y_test))#预测结果和真实标签
#准确率与召回率
precision, recall, thresholds = precision_recall_curve(y_test, clf.predict(x_test))
answer = clf.predict_proba(x_test)[:,1]
report = answer > 0.5
print(classification_report(y_test, report, target_names = ['neg', 'pos']))
print("--------------------")
from sklearn.metrics import accuracy_score
print('准确率: %.2f' % accuracy_score(y_test, doc_class_predicted))
if __name__ == '__main__':
data=[]
labels=[]
with open ("train2.txt","r")as file:
for line in file:
line=line[0:1]
labels.append(line)
with open("train2.txt","r")as file:
for line in file:
line=line[1:]
data.append(line)
x=np.array(data)
labels=np.array(labels)
labels=[int (i)for i in labels]
movie_target=labels
#转换成空间向量
count_vec = TfidfVectorizer(binary = False)
#加载数据集，切分数据集80%训练，20%测试
x_train, x_test, y_train, y_test= train_test_split(x, movie_target, test_size = 0.2)
x_train = count_vec.fit_transform(x_train)
x_test = count_vec.transform(x_test)
print('**************支持向量机************ ')
Precision(SvmClass(x_train, y_train))
print('**************朴素贝叶斯************ ')
Precision(NbClass(x_train, y_train))
print('**************最近邻ＫＮＮ************ ')
Precision(KnnClass(x_train,y_train))
print('**************逻辑回归************ ')
Precision(LogisticClass(x_train, y_train))
print('**************决策树************ ')
Precision(DccisionClass(x_train,y_train))
print('**************逻辑森林************ ')
Precision(random_forest_class(x_train,y_train))
结果如下：
对于整体代码和语料的下载，可以去下载
