Chatbots are AI-driven applications designed to engage in conversational interactions with users through text messaging. They provide valuable information and perform tasks that traditionally require significant human labor. Across diverse fields such as education, healthcare, entertainment, and commerce, chatbots play a pivotal role in enhancing user experiences. This paper explores the development of a specialized chatbot for Morocco’s tourism sector using state-of-the-art Large Language Models (LLMs). The chatbot integrates advanced natural language processing techniques, including fine-tuning and Retrieval-Augmented Generation (RAG), to provide personalized and informative assistance to travelers. Through meticulous data collection and preprocessing, we curated a comprehensive dataset tailored to tourism-related queries. Evaluation metrics such as BLEU, Rouge, and perplexity demonstrate the effectiveness of our approach, highlighting RAG’s ability to enhance response quality by leveraging external knowledge retrieval. This research contributes to advancing AI-driven solutions in tourism, aiming to optimize visitor satisfaction and operational efficiency in diverse cultural contexts.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Enhancing Tourism Experiences in Morocco: LLM-A Comparative Study of Fine-Tuning and Retrieval Augmented Generation with Llama-3

  • Laila Machkour,
  • Meriem El Brahimi,
  • Rajae Ben Lahmar,
  • Bouchra Bouhamidi,
  • El Habib Benlahmar,
  • Omar Zahour,
  • Brahim Zahour

摘要

Chatbots are AI-driven applications designed to engage in conversational interactions with users through text messaging. They provide valuable information and perform tasks that traditionally require significant human labor. Across diverse fields such as education, healthcare, entertainment, and commerce, chatbots play a pivotal role in enhancing user experiences. This paper explores the development of a specialized chatbot for Morocco’s tourism sector using state-of-the-art Large Language Models (LLMs). The chatbot integrates advanced natural language processing techniques, including fine-tuning and Retrieval-Augmented Generation (RAG), to provide personalized and informative assistance to travelers. Through meticulous data collection and preprocessing, we curated a comprehensive dataset tailored to tourism-related queries. Evaluation metrics such as BLEU, Rouge, and perplexity demonstrate the effectiveness of our approach, highlighting RAG’s ability to enhance response quality by leveraging external knowledge retrieval. This research contributes to advancing AI-driven solutions in tourism, aiming to optimize visitor satisfaction and operational efficiency in diverse cultural contexts.