Create your Gitee Account
Explore and code with more than 5 million developers,Free private repositories !:)
Sign up
This repository doesn't specify license. Without author's permission, this code is only for learning and cannot be used for other purposes.
Clone or download
Cancel
Notice: Creating folder will generate an empty file .keep, because not support in Git
Loading...
README.md

1、预处理

(1)、特征提取

  • 对应文件:feature_extraction.py

最后结果:

chi1

X^2值前几名的词语。能看出这些词都是一些有效的情感词。“了”这样的词出现在其中,说明可以去除一些停用词,来进一步提高分类精度。

chi2

X^2值后几名的词语。能看出这些词的分类作用不是很大。

(2)、结果评价

  • 对应文件:tools.py

结果展示

evaluation

2、基于情感词典的情感极性分析

—— sentiment analysis based on sentiment dict

  • 对应文件:classifier.py DictClassifier

使用1:analyse_sentence

analyse_sentence(sentence, runout_filepath=None, print_show=False)

对单个句子进行情感极性分析

  • sentence,待分析的句子

  • 若runout_filepath指定,则将分析结果写入该文件;

  • 若print_show为True,则在控制台输出分析结果。

运行实例:

d = DictClassifier()
a_sentence = "剁椒鸡蛋好咸,土豆丝很好吃"
result = ds.analyse_sentence(a_sentence)
print(result)

使用2:analysis_file

analysis_file(filepath_in, filepath_out, encoding="utf-8", print_show=False, start=0, end=-1)

  • filepath_in,待分析的句子文件

  • filepath_out,分析结果输出文件

  • encoding,输入文件字符编码

  • print_show,是否在控制台输出

  • start,输入文件开始分析的句子行数

  • end,输入文件结束分析的句子行数

输出实例:

送餐快,态度好!味道不错。
Score:6.0
Sub-clause0: positive:快 
Sub-clause1: positive:好 punctuation:! 
Sub-clause2: positive:不错 

还可以,比预计时间晚了一小时到,不过还好
Score:-0.56
Sub-clause0: positive:还可以 
Sub-clause1: negative:晚……小时:晚了一小时 小时 
Sub-clause2: conjunction:不过 positive:还好

3、基于k-NN的情感极性分析

—— sentiment analysis based on k-NN

single_k_classify(input_data)

使用单个k值

k = 3

knn = KNNClassifier(train_data, train_labels, k=2, best_words=best_words)
classify_labels = []

print("KNNClassifiers is testing ...")
for data in self.test_data:
    classify_labels.append(knn.classify(data))
print("KNNClassifiers tests over.")

filepath = "f_runout/KNN-train-%d-test-%d-k-%s-%s.xls" % \
           (train_num, test_num, k,
            datetime.datetime.now().strftime("%Y-%m-%d-%H-%M-%S"))

results = get_accuracy(test_labels, classify_labels)
Write2File.write_contents(filepath, results)

multiple_k_classify(input_data)

使用多个k值

from spa.classifiers import KNNClassifier

k = [1, 3, 5, 7, 9, 11, 13]

knn = KNNClassifier(train_data, train_labels, k=2, best_words=best_words)
classify_labels = []

print("KNNClassifiers is testing ...")
for data in self.test_data:
    classify_labels.append(knn.classify(data))
print("KNNClassifiers tests over.")

filepath = "f_runout/KNN-train-%d-test-%d-k-%s-%s.xls" % \
           (train_num, test_num, '-'.join([str(i) for i in k]),
            datetime.datetime.now().strftime("%Y-%m-%d-%H-%M-%S"))

results = get_accuracy(test_labels, classify_labels)
Write2File.write_contents(filepath, results)

比较结论

在某些特定数据下,multiple_k比每个single_k效果要好。但并不是总是最好。

4、基于Bayes的情感极性分析

—— sentiment analysis based on bayes

from spa.classifiers import BayesClassifier

bayes = BayesClassifier(self.train_data, self.train_labels, self.best_words)

classify_labels = []
print("BayesClassifier is testing ...")
for data in self.test_data:
    classify_labels.append(bayes.classify(data))
print("BayesClassifier tests over.")

filepath = "f_runout/bayes-train-%d-test-%d-k-%s-%s.xls" % \
           (train_num, test_num, '-'.join([str(i) for i in k]),
            datetime.datetime.now().strftime("%Y-%m-%d-%H-%M-%S"))

results = get_accuracy(test_labels, classify_labels)
Write2File.write_contents(filepath, results)

5、基于最大熵的情感极性分析

—— sentiment analysis based on maximum entropy

使用1:得到每次迭代的准确率的变化

maxent_iteration

def test_maxent_iteration(self):
    print("MaxEntClassifier iteration")
    print("---" * 45)
    print("Train num = %s" % self.train_num)
    print("Test num = %s" % self.test_num)
    print("maxiter = %s" % self.max_iter)

    from spa.classifiers import MaxEntClassifier

    m = MaxEntClassifier(self.max_iter)
    iter_results = m.test(self.train_data, self.train_labels, self.best_words, self.test_data)

    filepath = "f_runout/MaxEnt-iteration-%s-train-%d-test-%d-f-%d-maxiter-%d-%s.xls" % \
               (self.type,
                self.train_num,
                self.test_num,
                self.feature_num,
                self.max_iter,
                datetime.datetime.now().strftime(
                    "%Y-%m-%d-%H-%M-%S"))

    results = []
    for i in range(len(iter_results)):
        try:
            results.append(get_accuracy(self.test_labels, iter_results[i], self.parameters))
        except ZeroDivisionError:
            print("ZeroDivisionError")

    Write2File.write_contents(filepath, results)

使用2:单个句子的情感极性划分

def test_maxent(self):
    print("MaxEntClassifier")
    print("---" * 45)
    print("Train num = %s" % self.train_num)
    print("Test num = %s" % self.test_num)
    print("maxiter = %s" % self.max_iter)

    from spa.classifiers import MaxEntClassifier

    m = MaxEntClassifier(self.max_iter)
    m.train(self.train_data, self.train_labels, self.best_words)

    print("MaxEntClassifier is testing ...")
    classify_results = []
    for data in self.test_data:
        classify_results.append(m.classify(data))
    print("MaxEntClassifier tests over.")

    filepath = "f_runout/MaxEnt-%s-train-%d-test-%d-f-%d-maxiter-%d-%s.xls" % \
               (self.type,
                self.train_num, self.test_num,
                self.feature_num, self.max_iter,
                datetime.datetime.now().strftime(
                    "%Y-%m-%d-%H-%M-%S"))

    self.write(filepath, classify_results, 1)

6、基于SVM的情感极性分析

—— sentiment analysis based on SVM

依赖于scikit-learn库。准确率较高!

def test_svm(self):
    print("SVMClassifier")
    print("---" * 45)
    print("Train num = %s" % self.train_num)
    print("Test num = %s" % self.test_num)
    print("C = %s" % self.C)

    from spa.classifiers import SVMClassifier
    svm = SVMClassifier(self.train_data, self.train_labels, self.best_words, self.C)

    classify_labels = []
    print("SVMClassifier is testing ...")
    for data in self.test_data:
        classify_labels.append(svm.classify(data))
    print("SVMClassifier tests over.")

    filepath = "f_runout/SVM-%s-train-%d-test-%d-f-%d-C-%d-%s-lin.xls" % \
               (self.type,
                self.train_num, self.test_num,
                self.feature_num, self.C,
                datetime.datetime.now().strftime(
                    "%Y-%m-%d-%H-%M-%S"))

    self.write(filepath, classify_labels, 2)

7、几种情感分析方法比较

基于词典

  • 准确率:准确率较高(80%以上),随着人工工作量的增加,准确率增加

  • 优点:易于理解

  • 缺点:人工工作量大

基于k_NN

  • 准确率:很低(60% - 70%)

  • 优点:思想简单、算法简单

  • 缺点:准确率低;耗内存;耗时间

基于Bayes

  • 准确率:还可以(70% - 80%)

  • 优点:简单,高效,运算速度快,扩展性好

  • 缺点:准确率不高,达不到实用

基于最大熵

  • 准确率:比较高(83%以上)

  • 优点:准确率高

  • 缺点:训练时间久

基于SVM

  • 准确率:最高(85%以上)

  • 优点:准确率高

  • 缺点:训练耗时

Comments ( 0 )

Sign in for post a comment

Releases

No release

Contributors

All

Activities

load more
can not load any more