朴素贝叶斯实现文本分类

Posted 2021-04-25 阿华code

文本分类

词语去重¶

import randomimport jiebaimport sklearnfrom sklearn.naive_bayes import MultinomialNBimport numpy as npimport matplotlib.pyplot as pltimport pylab as pldef set_word(filename): word_set = set() with open(filename,'r+') as f: for line in f.readlines(): word = line.strip().decode('utf-8') if len(word) > 0 and word not in word_set: word_set.add(word) return word_set

文本处理

import osdef text_Processing(file_path,test_size = 0.2): file_list = os.listdir(file_path) data_list = [] class_list = [] for file in file_list: new_file_path = os.path.join(file_path,file) files = os.listdir(new_file_path) j = 1 for file_ in files: if j>100: break with open(os.join(new_file_path,file_),'r+') as f: raw = f.read() jieba.enable_parallel(4) #jieba并行分词（window不可用），参数为进程数 word_cut = jie.cut(raw,cut_all = False) #精确模式 words_list = list[word_cut] jieba.disable_parallel()#释放进程 data_list.append(words_list)  class_list.append(file.decode('utf-8')) #文件夹名就是分类 j += 1  ##划分训练集和测试集 data_class_list = zip(data_list,class_list) random.shuffle(data_class_list) index = int(len(data_class_list)*test_size) + 1 train_list = data_class_list[index:] test_list = data_class_listp[:index] train_data_list,train_class_list = zip(*train_list) test_data_list,test_class_list = zip(*test_list)  #统计词频 all_words_dict = {} for word_list in train_data_list: for word in word_list: if all_words_dict[word]: all_words_dict[word] += 1 else: all_words_dict[word] += 1 all_words_tuple_lsit = sorted(all_words_dict.items,key=lambda f:f[1],reverse=True) all_words_list = list(zip(*all_words_tuple_lsit)[0]) return all_words_list,train_data_list,test_class_list,train_class_list

选取特征词

def words_dict(all_words_list,deletN,stopwords_set = set()): feature_words = [] n = 1 for t in range(deletN,len(all_words_list),1): if n > 1000: #特征词的维度1000 break if not all_words_list[t].isdigit() and all_words_list[t] not in stopwords_set and 1 < len(all_words_list[t]) < 5: feature_words.append(all_words_list[t]) n += 1 return feature_words

文本特征

def text_features(train_data_list,test_data_list,feature_words,flag='sklearn'): def rapper(text,feature_words): text_words = set(text) if flag == 'nltk': features = {word:1 if word in text_words else 0 for word in feature_words} elif flag == 'sklearn': features = [1 if word in text_words else 0 for word in feature_words] else: features = [] return features train_feature_list = [rapper(text,feature_words) for text in train_data_list] test_feature_list = [rapper(text_,feature_words) for text_ in test_data_list] return train_feature_list,test_feature_list

分类，同时输出准确率

import nltkdef text_classifier(train_feature_list,test_feature_list,train_class_list,test_class_list,flag = 'sklearn'): if flag == 'nltk': train_flist = zip(train_feature_list,train_class_list) test_flist = zip(test_feature_list,test_class_list) classifier = nltk.classify.NaiveBayesClassifier.train(train_flist) test_accuracy = nltk.classify.accuracy(classifier,test_flist) elif flag == 'sklearn': classifier = MultinomialNB.fit(train_feature_list,train_class_list) #多项式朴素贝叶斯fit test_accuracy = classifier.score(test_feature_list,test_class_list) # 对测试集打分 else: test_accuracy = [] return test_accuracy

实战

print('start')##文本预处理file_path('./testdata')all_words_list,train_data_list,test_class_list,train_class_list = text_Processing(file_path,test_size=0.2)stopwords_file = './stopwords.txt'stopwords_set = set_word(stopwords_file)
## 文本特征提取和分类flag = 'sklearn'deleteNs = range(0,1000,20)test_accuracy_list = []for deleteN in deleteNs: feature_word = words_dict(all_words_list,deletN,stopwords_set) train_feature_list,test_feature_list = text_features(train_data_list,test_data_list,feature_word,flag) test_acctacy = text_classifier(train_feature_list,test_feature_list,train_class_list,test_class_list) test_accuracy_list.append(test_acctacy)print(test_accuracy_list)print(finished)

可以从搜狗语料库下载数据集