Leveraging Chatbot Model for Tourism Information Services using NLP and ANN

Abstract

The digital transformation of the tourism sector has increased the need for real-time and interactive information services. Many Community-Based Tourism (CBT) destinations in rural areas still depend on conventional information delivery methods, which limit accessibility and responsiveness. This paper proposes an Artificial Intelligence–based chatbot that utilizes Natural Language Processing and Artificial Neural Network techniques to improve tourism information services in CBT destinations. This study applies a case study approach in Kampung Gedong Village, a community-based tourism destination with strong historical and cultural value. We constructed the chatbot dataset through field observations and interviews with local stakeholders and defined 24 tourism-related intents covering attractions, cultural activities, facilities, accessibility, local etiquette, and community services. We applied standard text preprocessing steps and represented textual features using the Bag of Words method. We implemented intent classification using a multilayer perceptron ANN model. The experimental results show that the proposed chatbot achieved an intent classification accuracy of approximately 92% and delivered fast, consistent, and context-aware responses. These findings confirm that AI-driven conversational systems are technically feasible and effective for enhancing information services in rural CBT destinations when aligned with local cultural contexts. This study contributes empirical evidence to smart tourism research and presents a scalable chatbot model that can be adopted by similar community-based tourism destinations.