General Culture Subtask Results
Evaluating Arabic cultural understanding across diverse topics
| Rank | Team | Affiliation | Accuracy | Members |
|---|---|---|---|---|
|
1st
|
Human-centred AI (HAI) research group | ADAPT, MTU | 72.15% | Shehenaz Hossain, Haithem Afli, Mohammmed Hasanuzzaman |
|
2nd
|
Pulkit Chatwal | Institute of Petroleum Technology | 71.65% | Pulkit Chatwal |
|
3rd
|
AYA_Team | QCRI | 71.45% | Jannatul tajrin, Bir Ballav Roy |
| 4th | Phoenix | Mohammed VI Polytechnic University | 71.35% | Houdaifa atou, AtouIssam Ait Yahia, Ismail Berrada |
| 5th | CultranAI | HBKU | 70.50% | Youssef Ahmed, Hunzalah Hassan, Firoj Alam |
| 6th | ISL-NLP | Arab Academy for Science and Technology (AAST) | 67.60% | Mohamed Gomaa, Nour Eldian |
| 7th | Rafiul Biswas | Hamad Bin Khalifa University | 67.55% | Shimaa Ibrahim, Kais Attia, Rafiul Biswas, Mabrouka Bessghaier, Wajdi Zaghouani |
| 8th | Hamyaria | Hadhramout University, Cairo University | 65.90% | Walid Al-Dhabyani, Hamzah A. Alsayadi |
| 9th | Star | Alexandria University | 64.05% | Eman Elrefai, Esraa Fouad, Hassan Ehab |
Competition Statistics
General Culture Subtask
Participating Teams: 9
Best Accuracy: 72.15%
Average Accuracy: 68.57%
Islamic Culture Subtask
Participating Teams: 6
Best Accuracy: 84.22%
Average Accuracy: 76.39%
Test Datasets Available
The test splits used in this evaluation are now publicly available on Hugging Face: