Ground-Truth Subgraphs for Better Training and Evaluation of Knowledge Graph Augmented LLMs
Alberto Cattaneo, Carlo Luschi, Daniel Justus · 2025
Read Ground-Truth Subgraphs for Better Training and Evaluation of Knowledge Graph Augmented LLMs by Alberto Cattaneo, Carlo Luschi, Daniel Justus (2025) 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-2511-04473.txt.
Description
Alberto Cattaneo and two colleagues build ground-truth data for knowledge graph question answering. Language models grounded in knowledge graphs promise more factual answers. Yet few datasets reveal the exact facts a system should retrieve, so methods are hard to compare.
Their SynthKGQA framework generates questions from any knowledge graph with full ground-truth fact sets. Applied to a Wikidata-derived graph and checked against a current Wikidata dump, it yields the GTSQA dataset. The set is built to test zero-shot skill on unseen material. With known targets, retrieval methods can finally be scored on what they fetch, not just final answers.
Catalog record
| CREATOR | Alberto Cattaneo, Carlo Luschi, Daniel Justus |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 86 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-04473