Hi, I want to inject my schema model into neo4j graph.
Converting the schema into a cypher query and injecting it into the graph consumes too much LLM resources, so I want to improve this.
Is there a neo4j tool that can create a graph based on schema? Or even better if there is an API provided.
schema model example :
"DINING": [
{
"template_file": "dining_template.csv",
"used_llm": "llama3.3_70b",
"schema_aggregation": [
{
"title": "Resort",
"type": "object",
"description": "Node",
"properties": [
{
"name": "name",
"column_name": "resort",
"type": "string",
"description": "The name of the resort",
"examples": [
"Dolphin"
]
}
]
},
{
"title": "Restaurant",
"type": "object",
"description": "Node",
"properties": [
{
"name": "name",
"column_name": "restaurant_name",
"type": "string",
"description": "The name of the restaurant",
"examples": [
"Shula's Steak House"
]
},
{
"name": "category",
"column_name": "dining_category",
"type": "string",
"description": "The category of the restaurant",
"examples": [
"Signature"
]
},
{
"name": "operatingHours",
"column_name": "operating_hours",
"type": "string",
"description": "The operating hours of the restaurant",
"examples": [
"Daily 5:00pm - 11:00pm"
]
},
{
"name": "dressCode",
"column_name": "dress_code",
"type": "array",
"description": "The dress code of the restaurant",
"examples": [
[
"dress code1",
"dress code2"
]
],
"items": {
"type": "string"
}
}
]
},
{
"title": "HAS_RESTAURANT",
"type": "object",
"description": "Relationship",
"properties": [
{
"name": "_from",
"column_name": "resort",
"type": "string",
"description": "The resort",
"examples": [
"Resort"
]
},
{
"name": "_to",
"column_name": "restaurant_name",
"type": "string",
"description": "The restaurants",
"examples": [
"Restaurant"
]
}
]
}
],
"nodes": [
"Resort",
"Restaurant"
],
"relationships": [
"HAS_RESTAURANT"
],
"obtained_time": "2025-02-04 07:32:25"
}
],