Knowledge Graph - NL To Graph Query Translator

Translate natural language questions into Cypher or SPARQL queries for graph databases and knowledge graphs.

Install

openclaw skills install nl-to-graph-query-translator

Natural Language → Cypher / SPARQL Translator

Convert natural language questions into executable graph queries.

This skill translates human-readable requests into graph query languages such as Cypher (Neo4j) or SPARQL (RDF), allowing developers to interact with graph databases using natural language instead of writing complex query syntax.

Quick Start

Use Cases

Use this skill when a user wants to:

✅ Convert natural language questions into graph queries
✅ Generate Cypher queries for property graph databases
✅ Generate SPARQL queries for RDF triple stores
✅ Explore graph datasets using plain English
✅ Quickly prototype graph queries

Example requests:

"Find employees who work at Acme."
"Show all companies located in California."
"List people connected to Alice within two hops."
"Find products purchased by customers in London."

What This Skill Produces

The skill generates executable graph queries in supported query languages:

🔗 Cypher queries (Neo4j)
🔗 SPARQL queries (RDF triple stores)
📝 Query explanations
🎯 Query parameter templates

Translation Process

The translator follows a 5-step process to convert natural language into executable queries:

Identify Entities — Detect entity names (people, organizations, locations)
Identify Relationships — Extract relationship verbs (works at, located in, purchased, etc.)
Determine Query Type — Classify request type (nodes, relationships, paths, counts, aggregations)
Construct Query Pattern — Build graph patterns representing relationships
Generate Query Syntax — Convert patterns into Cypher or SPARQL

For detailed information about each step, see Query Translation Strategy.

Supported Query Languages

Cypher

Used in Neo4j and other property graph databases.

Example:

MATCH (p:Person)-[:PURCHASED]->(product:Product)
RETURN p, product

📖 See Cypher Query Guide

SPARQL

Used in RDF triple stores and knowledge graphs.

Example:

SELECT ?person ?product
WHERE {
  ?person :purchased ?product .
}

📖 See SPARQL Query Guide

Common Patterns

Simple Entity Lookup

"Find all people named Alice"

Relationship Traversal

"Show companies where employees work"

Multi-Hop Queries

"Find people connected to Alice within two relationships"

For more patterns, see Query Patterns Reference.

Examples

Browse practical examples organized by complexity and use case:

Best Practices

When generating and executing graph queries:

✓ Use schema labels consistently
✓ Avoid overly broad relationship patterns
✓ Prefer parameterized queries for security
✓ Include LIMIT clauses for large datasets
✓ Validate query syntax before execution

See Architecture & Implementation for advanced considerations.

Advanced Topics

Related Skills

Last Updated: 2026-03-08
Maintainer: OpenClaw Community
License: See LICENSE in repository root