[Bug] [Knowledge] Query: CALL db.vertexVectorKnnSearch('entity','_embedding', XXX使用, {top_k:None}) YIELD node WHERE node.distance < None RETURN node.id AS id;

[Hec-gitHub]

Search before asking

• I had searched in the issues and found no similar issues.

Operating system information

Linux

Python version information

3.10

DB-GPT version

latest release

Related scenes

• Chat Data
• Chat Excel
• Chat DB
• Chat Knowledge
• Model Management
• Dashboard
• Plugins

Installation Information

• Installation From Source

• Docker Installation

• Docker Compose Installation

• Cluster Installation

• AutoDL Image

• Other

Device information

CPU

Models information

Proxy LLM

What happened

创建完知识图谱后(Knowledge Graph)，与知识图谱对话时，返回错误。
File "/dbgpt/rag/retriever/embedding.py", line 267, in _similarity_search_with_score
return await self._index_store.asimilar_search_with_scores(
File "/dbgpt/storage/knowledge_graph/knowledge_graph.py", line 224, in asimilar_search_with_scores
subgraph = self._graph_store_adapter.explore_trigraph(
File "/dbgpt/storage/knowledge_graph/community/tugraph_store_adapter.py", line 672, in explore_trigraph
self.graph_store.conn.run(query=similarity_retrieval_query)
File "/dbgpt/datasource/conn_tugraph.py", line 127, in run
raise Exception(f"Query execution failed: {e}\nQuery: {query}") from e
Exception: Query execution failed: {code: CypherException} {message: Unknown variable: None}
Query: CALL db.vertexVectorKnnSearch('entity','_embedding', XXX使用, {top_k:None}) YIELD node WHERE node.distance < None RETURN node.id AS id;

            similarity_retrieval_query = (
                "CALL db.vertexVectorKnnSearch("
                f"'{GraphElemType.ENTITY.value}','_embedding', {vector}, "
                f"{{top_k:{topk}}}) YIELD node "
                f"WHERE node.distance < {score_threshold} "
                "RETURN node.id AS id;"
            )

手动将topk，score_threshold值固定，查询结果如下：
File "/dbgpt/storage/knowledge_graph/knowledge_graph.py", line 224, in asimilar_search_with_scores
subgraph = self._graph_store_adapter.explore_trigraph(
File "/dbgpt/storage/knowledge_graph/community/tugraph_store_adapter.py", line 674, in explore_trigraph
self.graph_store.conn.run(query=similarity_retrieval_query)
File "/dbgpt/datasource/conn_tugraph.py", line 127, in run
raise Exception(f"Query execution failed: {e}\nQuery: {query}") from e
Exception: Query execution failed: {code: CypherException} {message: Unknown variable: well}
Query: CALL db.vertexVectorKnnSearch('entity','_embedding', well, {top_k:5}) YIELD node WHERE node.distance < 0.3 RETURN node.id AS id;

另外，登录TuGraph web端，可以正常查询上述知识库内容。

What you expected to happen

正常返回查询结果

How to reproduce

知识图谱对话时，返回错误。

Additional context

No response

Are you willing to submit PR?

• Yes I am willing to submit a PR!

[Hec-gitHub]

知识图谱创建成功了，使用view graph 也可以看到图谱与数据正常显示。
另外，在TuGraph 客户端也看到了数据正常被创建，使用“entity”类型也可以正常查询数据。
(properties:{
name:"API使用Key",
id:"API使用Key",
description:null,
_community_id:null,
_embedding:null)。 "_community_id" 与 “_embedding” 的值都未创建。
嵌入模型使用的是：m3e-large

但是，使用Chat Knowledge 对话时，就会报查询错误。
错误信息：
File "/DB-GPT/dbgpt/storage/knowledge_graph/community/tugraph_store_adapter.py", line 684, in explore_trigraph
self.graph_store.conn.run(query=similarity_retrieval_query)

no viable alternative at input 'CALL db.vertexVectorKnnSearch('entity','_embedding', API service'} Query: CALL db.vertexVectorKnnSearch('entity','_embedding', API service, {top_k:None}) YIELD node WHERE node.distance < None RETURN node.id AS id;