from nltk.corpus import stopWords
from nltk.tokenize import word_tokenize
from nltk.stem import PorterStemmer
from nltk.classify import NaiveBayesClassifier

# 加载数据
data = [("我爱北京", "积极"), ("我讨厌北京", "消极")]

# 数据预处理
stop_words = set(stopwords.words("english"))
stemmer = PorterStemmer()
processed_data = []
for text, label in data:
tokens = word_tokenize(text)
filtered_tokens = [token for token in tokens if token not in stop_words]
stemmed_tokens = [stemmer.stem(token) for token in filtered_tokens]
processed_data.append((stemmed_tokens, label))

# 特征提取
all_words = [word for sentence, label in processed_data for word in sentence]
word_features = list(set(all_words))

def document_features(document):
document_words = set(document)
features = {}
for word in word_features:
features["contains({})".fORMat(word)] = (word in document_words)
return features

feature_sets = [(document_features(sentence), label) for sentence, label in processed_data]

# 模型训练
classifier = NaiveBayesClassifier.train(feature_sets)

# 模型评估
print(classifier.accuracy(feature_sets))