1
import os
2
import random
3
import pandas as pd
4
from nltk.tokenize import word_tokenize
5
from nltk.corpus import stopwords
6

7
# 读取数据集
8
def load_imdb_data(data_dir):
9
    texts = []
10
    labels = []
11
    # 遍历正面/负面评论文件夹
12
    for label in ['pos', 'neg']:
13
        folder_path = os.path.join(data_dir, label)
14
        for file in os.listdir(folder_path):
15
            with open(os.path.join(folder_path, file), 'r', encoding='utf-8') as f:
16
                texts.append(f.read())
17
                labels.append(1 if label == 'pos' else 0)  # 正面=1，负面=0
18
    # 打乱数据顺序
19
    combined = list(zip(texts, labels))
20
    random.shuffle(combined)
21
    return zip(*combined)  # 返回(texts, labels)
22

23
# 加载训练集（取前1000条作为演示，完整数据集需去掉切片）
24
train_texts, train_labels = load_imdb_data('aclImdb/train')
25
train_texts, train_labels = train_texts[:1000], train_labels[:1000]
26

27
# 文本预处理函数
28
def preprocess_text(text):
29
    # 分词
30
    tokens = word_tokenize(text.lower())
31
    # 去除停用词和非字母字符
32
    stop_words = set(stopwords.words('english'))
33
    tokens = [token for token in tokens if token.isalpha() and token not in stop_words]
34
    return ' '.join(tokens)  # 拼接成字符串供特征提取
35

36
# 预处理所有文本
37
processed_texts = [preprocess_text(text) for text in train_texts]

1
from sklearn.feature_extraction.text import TfidfVectorizer
2
from sklearn.naive_bayes import MultinomialNB
3
from sklearn.metrics import accuracy_score
4

5
# 将文本转换为数值特征（TF-IDF）
6
vectorizer = TfidfVectorizer(max_features=5000)  # 保留5000个最关键的词汇
7
X_train = vectorizer.fit_transform(processed_texts)
8
y_train = train_labels
9

10
# 训练朴素贝叶斯分类器
11
model = MultinomialNB()
12
model.fit(X_train, y_train)
13

14
# 简单测试
15
test_reviews = [
16
    "This movie was amazing! The acting was superb and the plot was gripping.",
17
    "Terrible film. I walked out after 10 minutes. Waste of money."
18
]
19
processed_tests = [preprocess_text(review) for review in test_reviews]
20
X_test = vectorizer.transform(processed_tests)
21
predictions = model.predict(X_test)
22

23
for review, pred in zip(test_reviews, predictions):
24
    print(f"评论: {review[:50]}...")
25
    print(f"预测情感: {'正面' if pred == 1 else '负面'}\n")