diff --git a/dbgpt/rag/knowledge/docx.py b/dbgpt/rag/knowledge/docx.py
index b10276347..7c1ecaa9f 100644
--- a/dbgpt/rag/knowledge/docx.py
+++ b/dbgpt/rag/knowledge/docx.py
@@ -2,6 +2,8 @@
 from typing import Any, Dict, List, Optional, Union
 
 import docx
+from docx.opc.oxml import parse_xml
+from docx.opc.pkgreader import _SerializedRelationship, _SerializedRelationships
 
 from dbgpt.core import Document
 from dbgpt.rag.knowledge.base import (
@@ -12,6 +14,21 @@ from dbgpt.rag.knowledge.base import (
 )
 
 
+def load_from_xml_v2(base_uri, rels_item_xml):
+    """Return |_SerializedRelationships| instance loaded with the relationships.
+
+    contained in *rels_item_xml*.collection if *rels_item_xml* is |None|.
+    """
+    srels = _SerializedRelationships()
+    if rels_item_xml is not None:
+        rels_elm = parse_xml(rels_item_xml)
+        for rel_elm in rels_elm.Relationship_lst:
+            if rel_elm.target_ref in ("../NULL", "NULL"):
+                continue
+            srels._srels.append(_SerializedRelationship(base_uri, rel_elm))
+    return srels
+
+
 class DocxKnowledge(Knowledge):
     """Docx Knowledge."""
 
@@ -47,8 +64,10 @@ class DocxKnowledge(Knowledge):
             documents = self._loader.load()
         else:
             docs = []
+            _SerializedRelationships.load_from_xml = load_from_xml_v2   # type: ignore
             doc = docx.Document(self._path)
             content = []
+
             for i in range(len(doc.paragraphs)):
                 para = doc.paragraphs[i]
                 text = para.text