Text2Cypher: Bridging Natural Language and Graph Databases
Makbule Gulcin Ozsoy, Leila Messallem, Jon Besga, Gianandrea Minneci · 2024
Read Text2Cypher: Bridging Natural Language and Graph Databases by Makbule Gulcin Ozsoy, Leila Messallem, Jon Besga, Gianandrea Minneci (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-10064.txt.
Description
Makbule Gulcin Ozsoy, Leila Messallem, Jon Besga, and Gianandrea Minneci build Text2Cypher, a tool that turns plain questions into graph database queries, in a paper from 2024. Knowledge graphs store facts as nodes and links, and the Cypher language reads them. Writing Cypher takes training that most users lack. Their tool aims to open graph data to non-experts.
Large language models can draft Cypher, but they often miss fine points. Fine-tuning on good examples works better, yet public training sets were scarce. The team merged and cleaned several sets into more than forty-four thousand examples. Models tuned on that pool gained a lot on match scores over plain models. The work shows data quality drives this task. It also gives the field a shared base for later gains.
Catalog record
| CREATOR | Makbule Gulcin Ozsoy, Leila Messallem, Jon Besga, Gianandrea Minneci |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-10064