Building a Text2cypher Model

Session Track: Data Science

Session Time: 09:30 pm - 09:45 pm November 7

Session description

Translating natural language into programming or domain-specific languages (DSL) is a common use of machine learning. A key example is the Text2Cypher task, where plain language queries are converted into Cypher query language. This can be driven by large language models (LLMs) or supervised models trained on datasets pairing natural language with Cypher translations. In this presentation, we'll cover Text2Cypher efforts at Neo4j, detailing the end-to-end process, from data preparation to model tuning and benchmarking. We'll also include a brief demo to showcase its practical application.