01Master's Thesis · Results being prepared for scientific publication
Intelligent System for Generating Arabic University Course Content from Real Data Using AI
Designed and developed an intelligent system for generating structured Arabic university lectures from real academic course and reference data using Large Language Models (LLMs). Fine-tuned and evaluated Qwen3-8B for domain-specific lecture generation, incorporating dataset development, data augmentation, prompt engineering, ontology-based knowledge representation, and multi-agent AI workflows across multiple processing stages. Conducted research-oriented experiments and comparative evaluation to improve generation quality, with results currently being prepared for scientific publication.
- 01DataReal course & reference data · dataset development · augmentation
- 02KnowledgeOntology-based knowledge representation
- 03ModelFine-tuned Qwen3-8B · prompt engineering
- 04AgentsMulti-agent AI workflows across processing stages
- 05EvaluateComparative evaluation · ROUGE · METEOR · BERTScore