Knowledge Graphs for RAG -- Query Knowledge Graphs (吴恩达-知识图谱在RAG中的应用 1) jupyter + langchain 版

Knowledge Graphs for RAG – Query Knowledge Graphs （吴恩达）

1导包并配置Neo4

from dotenv import load_dotenv
import os

from langchain_community.graphs import Neo4jGraph

# Warning control
import warnings
warnings.filterwarnings("ignore")

load_dotenv('.env', override=True)
NEO4J_URI = os.getenv('NEO4J_URI')#Neo4j的链接字符串，指明Neoc4j 运行于哪个端口
NEO4J_USERNAME = os.getenv('NEO4J_USERNAME')#用户名
NEO4J_PASSWORD = os.getenv('NEO4J_PASSWORD')#密码
NEO4J_DATABASE = os.getenv('NEO4J_DATABASE')#在DBMS中要使用的数据库名

1.1 使用LangChain的Neo4j初始化一个知识图谱实例

kg = Neo4jGraph(
    url=NEO4J_URI, username=NEO4J_USERNAME, password=NEO4J_PASSWORD, database=NEO4J_DATABASE
)

1.2 查询电影知识图谱

1.2.1 简单实例：人和电影的关系

1.2.2 匹配图形中的所有节点

# 查询图谱中的总节点数
cypher = """
  MATCH (n) 
  RETURN count(n)
  """

result = kg.query(cypher)
result

给 count(n) 新的命名使其显示更优雅

cypher = """
  MATCH (n) 
  RETURN count(n) AS numberOfNodes
  """

1.2.3 通过指定节点标签仅匹配电影节点

cypher = """
  MATCH (n:Movie) 
  RETURN count(n) AS numberOfMovies
  """
kg.query(cypher)

换个节点模式匹配中变量名增加可读性（movie 首字母是m，所以我们可以把n变量换成m）

1.2.4 通过在“person”节点上指定name属性的值来匹配单个个人

cypher = """
  MATCH (tom:Person {name:"Tom Hanks"}) 
  RETURN tom
  """
kg.query(cypher)

通过指定title属性的值来匹配单个Movie

cypher = """
  MATCH (cloudAtlas:Movie {title:"Cloud Atlas"}) 
  RETURN cloudAtlas
  """
kg.query(cypher)

如果不想返回整个cloudAtlas节点，而是只想要发行日（released）：

cypher = """
  MATCH (cloudAtlas:Movie {title:"Cloud Atlas"}) 
  RETURN cloudAtlas.released
  """
kg.query(cypher)

想要多个属性值时：

cypher = """
  MATCH (cloudAtlas:Movie {title:"Cloud Atlas"}) 
  RETURN cloudAtlas.released, cloudAtlas.tagline
  """
kg.query(cypher)

1.2.5 具有条件匹配的Cypher模式

cypher = """
  MATCH (nineties:Movie) 
  WHERE nineties.released >= 1990 
    AND nineties.released < 2000 
  RETURN nineties.title
  """

1.3 多节点+关系的模式匹配

cypher = """
  MATCH (actor:Person)-[:ACTED_IN]->(movie:Movie) 
  RETURN actor.name, movie.title LIMIT 10
  """
kg.query(cypher)

只查询某个名字的演员参演的电影及演员全名

cypher = """
  MATCH (tom:Person {name: "Tom Hanks"})-[:ACTED_IN]->(tomHanksMovies:Movie) 
  RETURN tom.name,tomHanksMovies.title
  """
kg.query(cypher)

查询与演员 ‘Tom Hanks’ 共同出演电影的演员名字及电影名字

# 两个node 变量： m 和 coActors， 先由Tom Hanks 查出m， 然后查出m 中的其他演员 coActors
cypher = """
  MATCH (tom:Person {name:"Tom Hanks"})-[:ACTED_IN]->(m)<-[:ACTED_IN]-(coActors) 
  RETURN coActors.name, m.title
  """
kg.query(cypher)

1.4 删除数据

# 查询Emil Eifrem 演过的电影及Emil Eifrem名字
cypher = """
MATCH (emil:Person {name:"Emil Eifrem"})-[actedIn:ACTED_IN]->(movie:Movie)
RETURN emil.name, movie.title
"""
kg.query(cypher)

# 删除 '演出' 关系
cypher = """
MATCH (emil:Person {name:"Emil Eifrem"})-[actedIn:ACTED_IN]->(movie:Movie)
DELETE actedIn
"""
kg.query(cypher)

1.5 添加数据

# 创建（CREATE） Person节点（andreas）
cypher = """
CREATE (andreas:Person {name:"Andreas"})
RETURN andreas
"""

kg.query(cypher)

#先查询出andreas 和emil ，然后创建（MERGE ）同事关系
cypher = """
MATCH (andreas:Person {name:"Andreas"}), (emil:Person {name:"Emil Eifrem"})
MERGE (andreas)-[hasRelationship:WORKS_WITH]->(emil)
RETURN andreas, hasRelationship, emil
"""
kg.query(cypher)