فاعلية هندسة التلقينات التوليدية في تطوير استجابات نماذج اللغة في السياقات البحثية العربية
This research aimed to improve academic writing by employing prompt engineering with generative artificial intelligence models such as ChatGPT and techniques including "Few shots" and "Chain of Thoughts". It explored the effectiveness of advanced prompt engineering techniques in enhancing the quality of large language models (LLMs), identified challenges in understanding and applying these technologies, and proposed a framework for integrating them into educational curricula. The research adopted descriptive and quasi-experimental methodologies through literature review followed by testing customized prompts with the participation of researchers and students who contributed feedback to refine the prompts. The impact of zero-shot prompting, few-shot prompting, chain prompting, and role-based prompting on response quality for educational tasks in Arabic was analyzed. Results indicated that advanced prompting techniques particularly chain prompting and role based prompting led to significant improvements in response accuracy, comprehensiveness, clarity, and relevance compared to zero-shot prompting. Statistically significant differences were found between the effectiveness of these techniques. The study also identified key challenges, including linguistic, cultural, ethical, training, and financial barriers. Practical guidelines were provided to enhance effective interaction with AI in education. The study recommended adopting advanced prompt engineering techniques by integrating their skills into training programs and effectively embedding AI in academic research to strengthen AI-supported scholarly writing in academic environments. (Published abstract)