上传文件至 /

上传文件
2025-04-03 17:23:53 +08:00
parent d0261c5997
commit c152fb8714
5 changed files with 1188 additions and 424 deletions
@@ -1,51 +1,55 @@
-"""
-===================================
-@Auther：WenZ
-@Company: BooWay
-@project：booway_dm
-===================================
-"""
-from vector_load import Mixed_retrieval, interface_search
-
-xizang_input_path = '../data/temp/Z1_keywords_software_usage.txt'
-retriever_txt_faiss1, retriever_txt_faiss2, retriever_txt_faiss3 = Mixed_retrieval(xizang_input_path)
-
-cuceng_input_path = '../data/temp/C1_keywords_software_usage.txt'
-retriever_txt_faiss4, retriever_txt_faiss5, retriever_txt_faiss6 = Mixed_retrieval(cuceng_input_path)
-
-jigai_input_path = '../data/temp/T1_keywords_software_usage.txt'
-retriever_txt_faiss7, retriever_txt_faiss8, retriever_txt_faiss9 = Mixed_retrieval(jigai_input_path)
-
-from vector_load import Building_search_dictionary, Official_website_kg_search
-
-input_index_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/data_index.csv"
-xizang_input_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/西藏造价FAQ数据集.csv"
-cuceng_input_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/新型储能电站建设计价通C1.csv"
-jigai_input_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/技改检修计价通T1.csv"
-
-
-def process_domain_category(nlu_info, retrievers, input_csv_path, index_csv_path):
-    """
-    处理不同领域类别的检索逻辑。
-
-    :param nlu_info: 包含领域类别和检索关键词的对象
-    :param retrievers: 用于检索的 FAISS 索引列表
-    :param input_csv_path: 领域特定的 CSV 输入路径
-    :param index_csv_path: 索引 CSV 路径
-    :return: 处理后的 QA RAG 结果列表
-    """
-    index_keywords = interface_search(nlu_info.retrieve_keywords, *retrievers)
-    qa_rag = []
-
-    for keyword in index_keywords:
-        # todo: bug修改: 避免output_id为None情况
-        output_path, output_id = Building_search_dictionary(input_csv_path, index_csv_path, keyword)
-        if output_path is not None or output_id is not None:
-            qa_rag.append(f"检索知识：{output_path}")
-            qa_rag.append(Official_website_kg_search(output_id))
-        else:
-            continue
-
-    return qa_rag
-
-
+"""
+===================================
+@Auther：WenZ
+@Company: BooWay
+@project：booway_dm
+===================================
+"""
+from vector_load import Mixed_retrieval, interface_search
+
+xizang_input_path = '../data/temp/Z1_keywords_software_usage.txt'
+retriever_txt_faiss1, retriever_txt_faiss2, retriever_txt_faiss3 = Mixed_retrieval(xizang_input_path)
+
+cuceng_input_path = '../data/temp/C1_keywords_software_usage.txt'
+retriever_txt_faiss4, retriever_txt_faiss5, retriever_txt_faiss6 = Mixed_retrieval(cuceng_input_path)
+
+jigai_input_path = '../data/temp/T1_keywords_software_usage.txt'
+retriever_txt_faiss7, retriever_txt_faiss8, retriever_txt_faiss9 = Mixed_retrieval(jigai_input_path)
+
+peiwang_input_path = '../data/temp/D3_keywords_software_usage.txt'
+retriever_txt_faiss10, retriever_txt_faiss11, retriever_txt_faiss12 = Mixed_retrieval(peiwang_input_path)
+
+from vector_load import Building_search_dictionary, Official_website_kg_search
+
+input_index_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/data_index.csv"
+xizang_input_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/西藏造价FAQ数据集.csv"
+cuceng_input_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/新型储能电站建设计价通C1.csv"
+jigai_input_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/技改检修计价通T1.csv"
+peiwang_input_csv_path = "/data/Z_project/rahulnyk/DM_lab/data/temp/博微配网工程计价通D3软件.csv"
+
+
+def process_domain_category(nlu_info, retrievers, input_csv_path, index_csv_path):
+    """
+    处理不同领域类别的检索逻辑。
+
+    :param nlu_info: 包含领域类别和检索关键词的对象
+    :param retrievers: 用于检索的 FAISS 索引列表
+    :param input_csv_path: 领域特定的 CSV 输入路径
+    :param index_csv_path: 索引 CSV 路径
+    :return: 处理后的 QA RAG 结果列表
+    """
+    index_keywords = interface_search(nlu_info.retrieve_keywords, *retrievers)
+    qa_rag = []
+
+    for keyword in index_keywords:
+        # todo: bug修改: 避免output_id为None情况
+        output_path, output_id = Building_search_dictionary(input_csv_path, index_csv_path, keyword)
+        if output_path is not None or output_id is not None:
+            qa_rag.append(f"检索知识：{output_path}")
+            qa_rag.append(Official_website_kg_search(output_id))
+        else:
+            continue
+
+    return qa_rag
+
+