rewordQALD9 is a bilingual benchmark with alternative rewordings of QALD questions, used for the task of Question Answering over Knowledge Graphs (KGQA). It aims to provide manually-curated translations of Italian questions from the QALD9 dataset. It is also used as a testbed for systems' robustness and flexibility in handing a wider range of language variations, possibly including both standard and informal register. In addition, a further enhancement of the resource is the addition of question reformulations to the original QALD9 data. The dataset can be download via link.
| Model / System | Year | F1-QALD | Language | Reported by |
|---|---|---|---|---|
| QAnswer | 2022 | 31.86 | EN | Sanguinetti et. al. |
| QAnswer | 2022 | 26.03 | IT | Sanguinetti et. al. |