Research-backed MCP system using Arabic morphology, transformer embeddings, and Poincaré geometry to ground semantic relationships in classical Arabic.
Arabic punctuation restoration project using NLP preprocessing, BiLSTM, CRF, FastText embeddings, and transformer-based token classification.
Arabic stance detection system for StanceEval-2026 on Mawqif-v2, combining fine-tuned Arabic encoders, retrieval-augmented few-shot LLMs, LoRA adapters, and a per-class stacking ensemble for seen and unseen target evaluation.
Bilingual (English + Arabic) fake review detector built with fine-tuned transformer models (DistilBERT + CamelBERT) and deployed as a FastAPI web app.
AI-powered tourism sentiment analysis platform for Arabic reviews and tourism analytics.
BAREC-ST-2026. The Second Shared Task on Sentence-level Readability Assessment (Open Track): An NLP competition on fine-grained Arabic readability classification across 19 levels.
To associate your repository with the camelbert topic, visit your repo's landing page and select "manage topics."