Open source, culturally aware LLMs for Arabic
Unit for Research in Arabic Social and Digital Spaces
Part of the Arab Center for Research and Policy Studies (ACRPS)
Advancing NLP, ML, and Computational Linguistics in Arabic digital spaces.
Projects
Automating editorial tasks for Arabic research
Chat with your favorite authors using AI
Custom URL shortener for ACRPS
Evaluating Arabic NLP and cultural awareness
Multimodal datasets across social spaces
WhatsApp agent for job search
Solution for live conference interactions
Publications
"R-BPE: Improving BPE-Tokenizers with Token Reuse"
EMNLP, 2025
"ImageEval 2025: The First Arabic Image Captioning Shared Task"
ArabicNLP Shared Tasks, 2025
"MASRAD: Arabic Terminology Management Corpora with Semi-Automatic Construction"
arXiv preprint, 2025
"AREEj: Arabic Relation Extraction with Evidence"
ArabicNLP, 2024
"DRU at WojoodNER 2024: A Multi-level Method Approach"
ArabicNLP, 2024
"Back-of-the-Book Index Automation for Arabic Documents"
arXiv preprint, 2024
"Arabic Topic Classification in the Generative and AutoML Era"
ArabicNLP, 2023
Resources
Relation extraction with evidence
With Wojood and Camel
With BERTopic
Shifts in research discourse on Palestine