正则表达式基础

1import re
2import pandas as pd
3
4df = pd.read_csv('reviews.csv')
5def tokenize_text(text):
6    return re.findall(r'\w+', text)
7df['tokens'] = df['review_text'].apply(tokenize_text)
8print(df['tokens'])

1import re
2import pandas as pd
3
4df = pd.read_csv('reviews.csv')
5def clean_text(text):
6    cleaned_text = re.sub(' +', ' ', re.sub('[^0-9a-zA-Z]+', ' ', text.lower()))
7    return cleaned_text
8df['cleaned_text'] = df['review_text'].apply(clean_text)
9print(df['cleaned_text'])

1import re
2import pandas as pd
3
4df = pd.read_csv('reviews.csv')
5patterns = {
6    'PERSON': r'(Barack Obama|John Doe|Jane Smith)',
7    'LOCATION': r'(Hawaii|United States|New York)'
8}
9named_entities = {}
10for entity, pattern in patterns.items():
11    def find_entities(text):
12        return re.findall(pattern, text)
13    df[entity] = df['review_text'].apply(find_entities)
14    named_entities[entity] = df[entity].tolist()
15print(named_entities)