Update predictor.py

predictor.py

@@ -149,6 +149,29 @@ class SentenceExtractor:
                 sentences.append(s)
         return sentences
     
+    def _fuzzy_match_keyword(self, sentence: str, keyword: str) -> bool:
+        """
+        模糊匹配关键词，支持部分匹配和相似词匹配
+        """
+        # 直接包含匹配
+        if keyword in sentence:
+            return True
+        
+        # 去除标点符号后匹配
+        import string
+        sentence_clean = sentence.translate(str.maketrans('', '', string.punctuation))
+        keyword_clean = keyword.translate(str.maketrans('', '', string.punctuation))
+        if keyword_clean in sentence_clean:
+            return True
+        
+        # 支持关键词的部分匹配（至少3个字符）
+        if len(keyword) >= 3:
+            for i in range(len(sentence) - len(keyword) + 1):
+                if sentence[i:i+len(keyword)] == keyword:
+                    return True
+        
+        return False
+
     def _extract_relevant_sentences(self, text: str) -> Tuple[List[str], Dict[str, List[str]]]:
         """
         提取与关键词相关的句子
@@ -156,8 +179,7 @@ class SentenceExtractor:
         :return: 相关句子列表和按类别分组的句子字典
         """
         sentences = self._split_into_sentences(text)
-        # 默认对所有分句进行评分，避免只命中少量关键词时数量过少
-        relevant_sentences = list(sentences)
+        relevant_sentences = []
         categorized_sentences = {
             "main": [],
             "student_performance": {"positive": [], "negative": [], "nature": [], "suggestion": []},
@@ -166,14 +188,22 @@ class SentenceExtractor:
         }
         
         for sentence in sentences:
-            # 检查是否包含主关键词（仅用于分类展示，不再决定是否计入评分）
+            sentence_added = False
+            
+            # 检查是否包含主关键词（使用模糊匹配）
             for category, keywords in self.main_keywords.items():
-                if any(keyword in sentence for keyword in keywords):
-                    if sentence not in categorized_sentences["main"]:
-                        categorized_sentences["main"].append(sentence)
+                for keyword in keywords:
+                    if self._fuzzy_match_keyword(sentence, keyword):
+                        if not sentence_added:
+                            relevant_sentences.append(sentence)
+                            sentence_added = True
+                        if sentence not in categorized_sentences["main"]:
+                            categorized_sentences["main"].append(sentence)
+                        break
+                if sentence_added:
                     break
 
-            # 检查评估关键词库中的关键词
+            # 检查评估关键词库中的关键词（使用模糊匹配）
             for category in ["student_performance", "content_quality", "cross_scene"]:
                 if category not in self.eval_keywords:
                     continue
@@ -183,8 +213,12 @@ class SentenceExtractor:
                         continue
                         
                     for keyword in self.eval_keywords[category][sentiment]:
-                        if keyword in sentence and sentence not in categorized_sentences[category][sentiment]:
-                            categorized_sentences[category][sentiment].append(sentence)
+                        if self._fuzzy_match_keyword(sentence, keyword):
+                            if not sentence_added:
+                                relevant_sentences.append(sentence)
+                                sentence_added = True
+                            if sentence not in categorized_sentences[category][sentiment]:
+                                categorized_sentences[category][sentiment].append(sentence)
         
         return relevant_sentences, categorized_sentences