'commit'

dsRag/ElasticSearch/T6_XiangLiangQuery.py

@@ -16,7 +16,8 @@ warnings.filterwarnings('ignore', message='Unverified HTTPS request is being mad
 
 if __name__ == "__main__":
     # 测试查询
-    query = "小学数学中有哪些模型"
+    # query = "小学数学中有哪些模型"
+    query = "文言虚词"
     query_tags = ["MATH_1"]  # 默认搜索标签，可修改
     print(f"\n=== 开始执行查询 ===")
     print(f"原始查询文本: {query}")
@@ -89,11 +90,14 @@ if __name__ == "__main__":
 
         # 打印详细结果
         print("\n=== 最终搜索结果 ===")
-        print(f" 向量搜索结果: {len(vector_results['hits']['hits'])}条")
+
+        vector_int = 0
         for i, hit in enumerate(vector_results['hits']['hits'], 1):
-            print(f"   {i}. 文档ID: {hit['_id']}, 相似度分数: {hit['_score']:.2f}")
-            print(f"      内容: {hit['_source']['user_input']}")
-            # print(f"      详细: {hit['_source']['tags']['full_content']}")
+            if hit['_score'] > 0.4: # 阀值0.4
+                print(f"   {i}. 文档ID: {hit['_id']}, 相似度分数: {hit['_score']:.2f}")
+                print(f"      内容: {hit['_source']['user_input']}")
+                vector_int = vector_int + 1
+        print(f" 向量搜索结果: {vector_int}条")
 
         print("\n文本精确搜索结果:")
         for i, hit in enumerate(text_results['hits']['hits']):

dsRag/Util/EsSearchUtil.py

@@ -172,7 +172,19 @@ class EsSearchUtil:
                     "size": 3
                 }
             )
-            logger.info(f"5. 向量搜索结果数量: {len(vector_results['hits']['hits'])}")
+            # 处理一下，判断是否到达阀值
+            filtered_vector_hits = []
+            vector_int = 0
+            for hit in vector_results['hits']['hits']:
+                if hit['_score'] > 0.4:  # 阀值0.4
+                    logger.info(f"   {vector_int + 1}. 文档ID: {hit['_id']}, 相似度分数: {hit['_score']:.2f}")
+                    logger.info(f"      内容: {hit['_source']['user_input']}")
+                    filtered_vector_hits.append(hit)
+                    vector_int += 1
+
+            # 更新vector_results只包含通过过滤的文档
+            vector_results['hits']['hits'] = filtered_vector_hits
+            logger.info(f"5. 向量搜索结果数量(过滤后): {vector_int}")
 
             # 文本精确搜索
             logger.info("\n=== 文本精确搜索阶段 ===")