Achieving Precise Text-To-Cypher Via Grounded Knowledge Graph Data Generation

2026-06-12

AI Summary

This work addresses the critical challenge of achieving high-accuracy natural language to Cypher query generation (Text-to-Cypher) in property graph databases while preserving data sovereignty. The authors propose a knowledge graph–guided approach for automatically synthesizing high-quality training data without manual annotation, which is then used to fine-tune small-scale large language models. This method substantially enhances both the accuracy and generalization capability of these compact models on Text-to-Cypher tasks, enabling them to match the performance of much larger, closed-source counterparts on standard benchmarks. Crucially, the approach supports local deployment, ensures data privacy, and maintains high query precision, offering a practical solution for real-world applications where data control and model efficiency are paramount.

Abstract

Property Graphs are rapidly being adopted as database frameworks for representing heterogeneous data sources. To enable precise access to the information contained in them we need conversational interfaces based on Text-To-Cypher (Text2Cypher) parsers. This paper presents an automatic synthetic data generation method that can be leveraged to fine-tune small LLMs for this task. We conduct experiments on all the major Text-To-Cypher benchmarks, demonstrating that with our synthetic data generation approach we can significantly increase the performance of small LLMs, allowing them to compete with much larger proprietary models. This means that in settings in which models must be locally deployed we can ensure data-sovereignty without sacrificing accuracy and without costly annotation campaigns.

Problem

• Text-To-Cypher
• Property Graphs
• Knowledge Graph
• Data Sovereignty
• LLM Fine-tuning

Innovation

• Text-to-Cypher
• synthetic data generation
• knowledge graph
• small LLMs
• data sovereignty