使用 HanaSparqlQAAgent 进行问答

设置与安装

要使用此功能，请安装 langchain-hana 包：

pip install langchain_hana

HanaSparqlQAAgent 是一种基于智能体的方法，用于回答存储在 SAP HANA Cloud 中的 RDF 数据的问题。与基于链的方法不同，智能体可以：

动态检索本体 使用专用工具
生成并执行 SPARQL 查询 迭代
自我纠正 如果查询失败或返回意外结果
逐步推理 关于复杂问题

初始化

您需要：

一个 LLM 为智能体的推理提供支持
一个 HanaRdfGraph（包含连接、graph_uri 和本体）

请按照此处 HanaRdfGraph 的步骤了解更多关于创建 HanaRdfGraph 实例的信息。导入 HanaSparqlQAAgent

from langchain_hana import HanaSparqlQAAgent
agent = HanaSparqlQAAgent.create_agent(graph=graph, model=llm)

智能体概述

智能体使用工具迭代执行以下操作：

首先检索本体以理解数据结构
根据模式生成适当的 SPARQL 查询
执行查询并解释结果
制定自然语言答案

默认设置

工具:
- retrieve_ontology - 从图中以 Turtle 格式检索 RDF 本体/模式
- execute_sparql - 对 HANA RDF 图执行 SPARQL 查询
系统提示词：提供默认系统提示词，包含 SPARQL 生成和工具使用的说明。
中间件：ModelRetryMiddleware(max_retries=3) 和 ToolRetryMiddleWare(max_retries=2) 默认添加以防止无限循环

自定义您的智能体

您可以通过向 create_agent 提供附加参数来自定义智能体的行为：

agent = HanaSparqlQAAgent.create_agent(
    graph=graph,  # Required: HanaRdfGraph instance
    model=llm,  # Required: LLM to power the agent
    system_prompt=your_custom_prompt,  # Custom system prompt (str or SystemMessage)
    tools=[your_custom_tools],  # Additional tools to include
    middleware=[your_custom_middlewares],  # Additional middleware to include
    include_default_tools=True,  # include default tools (default: True)
    include_default_middleware=True,  # include default middleware (default: True)
)

示例：在 “Movies” 知识图谱上进行问答

前提条件：您必须有一个启用了 三元组存储 功能的 SAP HANA Cloud 实例。有关详细说明，请参阅：启用三元组存储
加载 kgdocu_movies 示例数据。参见知识图谱示例。以下示例：

实例化 HanaRdfGraph 指向电影数据图
创建一个由 LLM 驱动的 HanaSparqlQAAgent
提问自然语言问题，并让智能体通过推理得出答案

这展示了智能体如何动态检索本体、生成 SPARQL 查询并返回人类可读的答案。

import os

from dotenv import load_dotenv
from hdbcli import dbapi

# Load environment variables if needed
load_dotenv()

# Establish connection to SAP HANA Cloud
connection = dbapi.connect(
    address=os.environ.get("HANA_DB_ADDRESS"),
    port=os.environ.get("HANA_DB_PORT"),
    user=os.environ.get("HANA_DB_USER"),
    password=os.environ.get("HANA_DB_PASSWORD")
)

然后，设置知识图谱实例

from gen_ai_hub.proxy.langchain.openai import ChatOpenAI
from langchain_hana import HanaRdfGraph, HanaSparqlQAChain

# from langchain_openai import ChatOpenAI  # or your chosen LLM

# Set up the Knowledge Graph
graph_uri = "kgdocu_movies"

graph = HanaRdfGraph(
    connection=connection,
    graph_uri=graph_uri,
    auto_extract_ontology=True
)

# a basic graph schema is extracted from the data graph. This schema will guide the LLM to generate a proper SPARQL query.
schema_graph = graph.get_schema
print(schema_graph.serialize(format="turtle"))

@prefix owl: <http://www.w3.org/2002/07/owl#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

<http://kg.demo.sap.com/acted_in> a owl:ObjectProperty ;
    rdfs:label "acted_in" ;
    rdfs:domain <http://kg.demo.sap.com/Actor> ;
    rdfs:range <http://kg.demo.sap.com/Film> .

<http://kg.demo.sap.com/dateOfBirth> a owl:DatatypeProperty ;
    rdfs:label "dateOfBirth" ;
    rdfs:domain <http://kg.demo.sap.com/Actor> ;
    rdfs:range xsd:dateTime .

<http://kg.demo.sap.com/directed> a owl:ObjectProperty ;
    rdfs:label "directed" ;
    rdfs:domain <http://kg.demo.sap.com/Director> ;
    rdfs:range <http://kg.demo.sap.com/Film> .

<http://kg.demo.sap.com/genre> a owl:ObjectProperty ;
    rdfs:label "genre" ;
    rdfs:domain <http://kg.demo.sap.com/Film> ;
    rdfs:range <http://kg.demo.sap.com/Genre> .

<http://kg.demo.sap.com/placeOfBirth> a owl:ObjectProperty ;
    rdfs:label "placeOfBirth" ;
    rdfs:domain <http://kg.demo.sap.com/Actor> ;
    rdfs:range <http://kg.demo.sap.com/Place> .

<http://kg.demo.sap.com/title> a owl:DatatypeProperty ;
    rdfs:label "title" ;
    rdfs:domain <http://kg.demo.sap.com/Film> ;
    rdfs:range xsd:string .

rdfs:label a owl:DatatypeProperty ;
    rdfs:label "label" ;
    rdfs:domain <http://kg.demo.sap.com/Actor>,
        <http://kg.demo.sap.com/Director>,
        <http://kg.demo.sap.com/Genre>,
        <http://kg.demo.sap.com/Place> ;
    rdfs:range xsd:string .

<http://kg.demo.sap.com/Director> a owl:Class ;
    rdfs:label "Director" .

<http://kg.demo.sap.com/Genre> a owl:Class ;
    rdfs:label "Genre" .

<http://kg.demo.sap.com/Place> a owl:Class ;
    rdfs:label "Place" .

<http://kg.demo.sap.com/Actor> a owl:Class ;
    rdfs:label "Actor" .

<http://kg.demo.sap.com/Film> a owl:Class ;
    rdfs:label "Film" .

之后，初始化 LLM。

# Initialize the LLM
llm = ChatOpenAI(proxy_model_name="gpt-4o", temperature=0)

然后，创建 SPARQL 问答智能体。

# Create a SPARQL QA Agent
agent = HanaSparqlQAAgent.create_agent(
    graph=graph,
    model=llm,
)

# Example queries:
# question = "Which movies are in the data?"
# question = "In which movies did Keanu Reeves and Carrie-Anne Moss play together"
# question = "which movie genres are in the data?"
# question = "which are the two most assigned movie genres?"
# question = "where were the actors of 'Blade Runner' born?"
# question = "which actors acted together in a movie and were born in the same city?"
question = "which actors acted in Blade Runner?"

for event in agent.stream(
    {"messages": [{"role": "user", "content": question}]},
    stream_mode="values",
):
    event["messages"][-1].pretty_print()

================================ Human Message =================================

which actors acted in Blade Runner?
================================== Ai Message ==================================
Tool Calls:
  retrieve_ontology (call_9T3Dxy3zKczTQKVAgTxPBfQm)
 Call ID: call_9T3Dxy3zKczTQKVAgTxPBfQm
  Args:
================================== Ai Message ==================================
Tool Calls:
  retrieve_ontology (call_9T3Dxy3zKczTQKVAgTxPBfQm)
 Call ID: call_9T3Dxy3zKczTQKVAgTxPBfQm
  Args:
================================= Tool Message =================================
Name: retrieve_ontology

Ontology Information:
@prefix owl: <http://www.w3.org/2002/07/owl#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

<http://kg.demo.sap.com/acted_in> a owl:ObjectProperty ;
    rdfs:label "acted_in" ;
    rdfs:domain <http://kg.demo.sap.com/Actor> ;
    rdfs:range <http://kg.demo.sap.com/Film> .

<http://kg.demo.sap.com/dateOfBirth> a owl:DatatypeProperty ;
    rdfs:label "dateOfBirth" ;
    rdfs:domain <http://kg.demo.sap.com/Actor> ;
    rdfs:range xsd:dateTime .

<http://kg.demo.sap.com/directed> a owl:ObjectProperty ;
    rdfs:label "directed" ;
    rdfs:domain <http://kg.demo.sap.com/Director> ;
    rdfs:range <http://kg.demo.sap.com/Film> .

<http://kg.demo.sap.com/genre> a owl:ObjectProperty ;
    rdfs:label "genre" ;
    rdfs:domain <http://kg.demo.sap.com/Film> ;
    rdfs:range <http://kg.demo.sap.com/Genre> .

<http://kg.demo.sap.com/placeOfBirth> a owl:ObjectProperty ;
    rdfs:label "placeOfBirth" ;
    rdfs:domain <http://kg.demo.sap.com/Actor> ;
    rdfs:range <http://kg.demo.sap.com/Place> .

<http://kg.demo.sap.com/title> a owl:DatatypeProperty ;
    rdfs:label "title" ;
    rdfs:domain <http://kg.demo.sap.com/Film> ;
    rdfs:range xsd:string .

rdfs:label a owl:DatatypeProperty ;
    rdfs:label "label" ;
    rdfs:domain <http://kg.demo.sap.com/Actor>,
        <http://kg.demo.sap.com/Director>,
        <http://kg.demo.sap.com/Genre>,
        <http://kg.demo.sap.com/Place> ;
    rdfs:range xsd:string .

<http://kg.demo.sap.com/Director> a owl:Class ;
    rdfs:label "Director" .

<http://kg.demo.sap.com/Genre> a owl:Class ;
    rdfs:label "Genre" .

<http://kg.demo.sap.com/Place> a owl:Class ;
    rdfs:label "Place" .

<http://kg.demo.sap.com/Actor> a owl:Class ;
    rdfs:label "Actor" .

<http://kg.demo.sap.com/Film> a owl:Class ;
    rdfs:label "Film" .


================================== Ai Message ==================================
Tool Calls:
  execute_sparql (call_LwHuDXaPsx3nDJF7Hn6If8s9)
 Call ID: call_LwHuDXaPsx3nDJF7Hn6If8s9
  Args:
    query: PREFIX kg: <http://kg.demo.sap.com/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT ?actorLabel
FROM <kgdocu_movies>
WHERE {
    ?movie rdf:type kg:Film .
    ?movie kg:title "Blade Runner" .
    ?actor kg:acted_in ?movie .
    ?actor rdfs:label ?actorLabel .
}
================================== Ai Message ==================================
Tool Calls:
  execute_sparql (call_LwHuDXaPsx3nDJF7Hn6If8s9)
 Call ID: call_LwHuDXaPsx3nDJF7Hn6If8s9
  Args:
    query: PREFIX kg: <http://kg.demo.sap.com/>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
SELECT ?actorLabel
FROM <kgdocu_movies>
WHERE {
    ?movie rdf:type kg:Film .
    ?movie kg:title "Blade Runner" .
    ?actor kg:acted_in ?movie .
    ?actor rdfs:label ?actorLabel .
}
================================= Tool Message =================================
Name: execute_sparql

SPARQL Query Result:
actorLabel
Morgan Paull
William Sanderson
James Hong
M. Emmet Walsh
Joe Turkel
Brion James
Q81328
Daryl Hannah
Rutger Hauer
Joanna Cassidy
Hy Pyke
Sean Young
Edward James Olmos

================================== Ai Message ==================================

The actors who acted in "Blade Runner" are:

- Morgan Paull
- William Sanderson
- James Hong
- M. Emmet Walsh
- Joe Turkel
- Brion James
- Daryl Hannah
- Rutger Hauer
- Joanna Cassidy
- Hy Pyke
- Sean Young
- Edward James Olmos

================================== Ai Message ==================================

The actors who acted in "Blade Runner" are:

- Morgan Paull
- William Sanderson
- James Hong
- M. Emmet Walsh
- Joe Turkel
- Brion James
- Daryl Hannah
- Rutger Hauer
- Joanna Cassidy
- Hy Pyke
- Sean Young
- Edward James Olmos

底层工作原理是什么？

本体检索 智能体首先使用 retrieve_ontology 工具以 Turtle 格式获取 RDF 模式。这使其了解可用的类、属性和关系。
SPARQL 生成 根据本体和用户的问题，智能体推理要查询的实体和属性，然后生成一个有效的 SELECT 查询。
查询执行 智能体调用 execute_sparql 工具在 HANA RDF 图上运行生成的查询。如果查询失败或返回意外结果，智能体可以自我纠正并重试。
答案制定 智能体解释查询结果，并严格基于检索到的数据制定简洁、人类可读的答案。

官方原文：src/oss/python/integrations/agents/sap_hana_sparql_qa_agent.mdx · 非官方翻译，译文由 AI 生成，以英文原文为准。

​设置与安装

​初始化

​智能体概述

​默认设置

​自定义您的智能体

​示例：在 “Movies” 知识图谱上进行问答

​底层工作原理是什么？