feat: 垃圾信息分类标签功能

新增垃圾信息细分类标签，在朴素贝叶斯二分类基础上对spam进行细分：
- 新增 spam_categorizer.py 分类模块（诈骗/骚扰/广告）
- SpamPredictionLog 和 ContentPost 模型添加 category 字段
- content_routes 和 spam_routes 接口返回分类标签

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/app/models.py

@@ -79,6 +79,7 @@ class SpamPredictionLog(db.Model):
     user_id = db.Column(db.Integer, db.ForeignKey("users.id"), nullable=False, index=True)
     text = db.Column(db.Text, nullable=False)
     prediction = db.Column(db.String(16), nullable=False)  # spam | ham
+    category = db.Column(db.String(32), default="")  # fraud | harassment | advertisement | spam | 空
     spam_probability = db.Column(db.Float, nullable=False)
     ham_probability = db.Column(db.Float, nullable=False)
     confidence = db.Column(db.Float, nullable=False)
@@ -92,6 +93,7 @@ class SpamPredictionLog(db.Model):
             "user_id": self.user_id,
             "text": self.text,
             "prediction": self.prediction,
+            "category": self.category or "",
             "spam_probability": round(float(self.spam_probability), 4),
             "ham_probability": round(float(self.ham_probability), 4),
             "confidence": round(float(self.confidence), 4),
@@ -130,6 +132,7 @@ class ContentPost(db.Model):
 
     status = db.Column(db.String(16), nullable=False, default="published")  # published | blocked
     prediction = db.Column(db.String(16), nullable=False, default="ham")
+    category = db.Column(db.String(32), default="")  # fraud | harassment | advertisement | spam | 空
     spam_probability = db.Column(db.Float, nullable=False, default=0)
     ham_probability = db.Column(db.Float, nullable=False, default=0)
     confidence = db.Column(db.Float, nullable=False, default=0)
@@ -163,6 +166,7 @@ class ContentPost(db.Model):
             "visibility": self.visibility,
             "status": self.status,
             "prediction": self.prediction,
+            "category": self.category or "",
             "spam_probability": round(float(self.spam_probability), 4),
             "ham_probability": round(float(self.ham_probability), 4),
             "confidence": round(float(self.confidence), 4),