campus_activity_system/utils/nlp_utils.py

from transformers import pipeline
import re
# 加载预训练的BERT分类器
classifier = pipeline("text-classification", model="distilbert-base-uncased-finetuned-sst-2-english")

def classify_activity_type(description):
    """根据描述文本对活动进行类型分类"""
    result = classifier(description)[0]
    label = result['label']
    
    # 简化标签映射到具体活动类型
    if 'positive' in label.lower():
        return "学术讲座"
    elif 'negative' in label.lower():
        return "文体比赛"
    else:
        return "其他"

ner_model = pipeline("ner", model="dslim/bert-base-NER")

def extract_information(text):
    """从文本中抽取姓名和身份证号"""
    result = ner_model(text)

    name = None
    id_number = None

    for entity in result:
        if entity['entity'] == 'B-PER':  # 姓名
            if name is None:
                name = entity['word'].replace('#', '')
        elif entity['entity'] == 'I-PER':
            name += entity['word'].replace('#', '')

        # 使用正则表达式匹配身份证号
        match = re.search(r'\d{17}[\dxX]|\d{15}', text)
        if match:
            id_number = match.group(0)

    return {'name': name, 'id_number': id_number}