diff --git a/ruoyi-fastapi-backend/utils/match_service.py b/ruoyi-fastapi-backend/utils/match_service.py
index e1e5627..7cf5716 100644
--- a/ruoyi-fastapi-backend/utils/match_service.py
+++ b/ruoyi-fastapi-backend/utils/match_service.py
@@ -78,7 +78,7 @@ class MatchService:
             '我们', '你们', '他们', '她们', '它们',
             '这个', '那个', '这些', '那些',
             '这种', '那种', '这类', '那类',
-            '你好', '您好', '大家好', '嗨', '哈喽', '哈罗',
+            '嗨', '哈喽', '哈罗',
             '早上好', '上午好', '下午好', '晚上好',
             'hi', 'hello', 'hey', '嗨', '哈喽',
             '怎么样', '如何', '怎样',
@@ -111,7 +111,18 @@ class MatchService:
             return ''
         
         text = text.strip()
+        
         text = re.sub(r'[^\u4e00-\u9fa5a-zA-Z0-9，。？！、；：""''（）【】《》\s]', '', text)
+        
+        colloquial_fillers = ['嗯', '呃', '额']
+        for filler in colloquial_fillers:
+            text = re.sub(f'{filler}+', filler, text)
+        
+        trailing_fillers = ['嗯', '哦', '啊', '呀', '啦', '呗', '嘛', '吧', '噢', '呃', '额', '哈']
+        for filler in trailing_fillers:
+            text = re.sub(f'{filler}\\s*$', '', text)
+            text = re.sub(f'{filler}$', '', text)
+        
         text = re.sub(r'\s+', ' ', text)
         
         for colloquial, standard in self._colloquial_map.items():
@@ -166,8 +177,8 @@ class MatchService:
         if n1 == n2:
             return 1.0
         
-        kw1 = set(self.extract_keywords(q1))
-        kw2 = set(self.extract_keywords(q2))
+        kw1 = set(self.extract_keywords(n1))
+        kw2 = set(self.extract_keywords(n2))
         
         if not kw1 or not kw2:
             return 0.0