Skip to content

LLM Prompts used in GraphRAG

Entity Extraction Prompts

Prompt #1

1
2
3
4
5
Please analyze the following text and extract the entities related to people,
organizations, locations, and events. For each entity, return the output in a
JSON format where each entity is represented as a node, and relationships
between entities are represented as edges. 
Ensure that the output follows this structure:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
{
    "nodes": [
        {"id": "1", "label": "Person", "name": "John Doe"},
        {"id": "2", "label": "Organization", "name": "ACME Corp"},
        {"id": "3", "label": "Location", "name": "New York"},
        {"id": "4", "label": "Event", "name": "Annual Conference 2024"}
    ],
    "edges": [
        {"source": "1", "target": "2", "relationship": "works_at"},
        {"source": "1", "target": "3", "relationship": "lives_in"},
        {"source": "2", "target": "4", "relationship": "organizes"},
        {"source": "3", "target": "4", "relationship": "location_of"}
    ]
}

1
Please make sure to include appropriate relationships between the entities wherever applicable.  Avoid duplication of edges.

Prompt #2

In this version, we only ask the LLM to extract only the most important entities and avoid duplication of relationships.

1
2
3
4
5
6
7
Please analyze the following text and extract the most important
entities related to people, organizations, locations, and events.
For each entity, return the output in a JSON format where each entity 
is represented as a node, and relationships between entities are
represented as edges.

Ensure that the output follows this structure:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
{
   "nodes": [
       {"id": "1", "label": "Person", "name": "John Doe"},
       {"id": "2", "label": "Organization", "name": "ACME Corp"},
       {"id": "3", "label": "Location", "name": "New York"},
       {"id": "4", "label": "Event", "name": "Annual Conference 2024"}
   ],
   "edges": [
       {"source": "1", "target": "2", "relationship": "works_at"},
       {"source": "1", "target": "3", "relationship": "lives_in"},
       {"source": "2", "target": "4", "relationship": "organizes"},
       {"source": "3", "target": "4", "relationship": "location_of"}
   ]
}

1
Please make sure to include appropriate relationships between the entities wherever applicable.  Avoid duplication of edges.