CrunchQA [1] is a KGQA dataset based on a knowledge graph derived from the Crunchbase, which contains information about companies ranging from early-stage startups to large companies. The questions are automatically generated from a set of carefully designed templates and paraphrases that capture several complexity levels. The knowledge graph CrunchKG includes 3.2 million entities, 31 relations, and 17.6 million triples.
| Model / System |
Year |
Hits@1 |
Language |
Reported by |
| TransE(clustering) [1] |
2023 |
15.6 |
EN |
Yu et al. |
| ComplEx(clustering) [1] |
2023 |
12.4 |
EN |
Yu et al. |
| DistMult(clustering) [1] |
2023 |
10.2 |
EN |
Yu et al. |
| Model / System |
Year |
Hits@1 |
Language |
Reported by |
| DistMult(clustering) [1] |
2023 |
8.9 |
EN |
Yu et al. |
| ComplEx(clustering) [1] |
2023 |
8.8 |
EN |
Yu et al. |
| TransE(clustering) [1] |
2023 |
8.1 |
EN |
Yu et al. |
| Model / System |
Year |
Hits@1 |
Language |
Reported by |
| TransE(clustering) [1] |
2023 |
37.4 |
EN |
Yu et al. |
| ComplEx(clustering) [1] |
2023 |
33.5 |
EN |
Yu et al. |
| DistMult(clustering) [1] |
2023 |
8.8 |
EN |
Yu et al. |
| Model / System |
Year |
Hits@1 |
Language |
Reported by |
| TransE(clustering) [1] |
2023 |
18.2 |
EN |
Yu et al. |
| TransE [1] |
2023 |
17 |
EN |
Yu et al. |
| ComplExLiteral [1] |
2023 |
13.8 |
EN |
Yu et al. |
| ComplEx(clustering) [1] |
2023 |
13.2 |
EN |
Yu et al. |
| DistMult [1] |
2023 |
12.7 |
EN |
Yu et al. |
| ComplEx [1] |
2023 |
12 |
EN |
Yu et al. |
| DistMult(clustering) [1] |
2023 |
11.1 |
EN |
Yu et al. |
| DistMultGatedLiteral [1] |
2023 |
10.6 |
EN |
Yu et al. |
- [1] Those methods are based on EmbedKGQA framework.
[1] Yu, Lifan and Madjid, Nadya Abdel and Difallah, Djellel. CrunchQA: A Synthetic Dataset for Question Answering over Crunchbase Knowledge Graph. In IEEE International Conference on Big Data (Big Data), pp. 4635-4641. 2022.
Go back to the README