A Smart Tourism and Culinary Itinerary Recommender Based on User Preferences and Time Constraints in Banyuwangi, Indonesia

Abstract

Tourism development in many regions requires intelligent systems capable of guiding visitors to destinations that match their preferences and available resources. This study proposes a novel Travel Recommendation System (TRS) capable of automatically generating integrated tourism and culinary itineraries while considering user preferences and time constraints. The proposed system incorporates several factors, including attraction categories, transportation modes, travel duration, budget limitations, and culinary preferences. A Content-Based Filtering approach is utilized to identify destinations that correspond to user characteristics and interests. Subsequently, a Genetic Algorithm (GA) is employed to optimize itinerary sequences, while a fuzzy logic mechanism is applied within the fitness evaluation process to assess budget suitability. Experimental results demonstrate that the integration of culinary destinations contributes to better itinerary quality compared to non-culinary itineraries, achieving an average fitness score difference of 0.018065. The proposed GA was compared with several other metaheuristic algorithms, showing superior performance in achieving higher fitness values and avoiding local optima. The generated itineraries were also evaluated by tourism experts, obtaining an average satisfaction score of 3.3. Overall, the proposed TRS shows great potential in producing practical, personalized, and high-quality itineraries that effectively integrate tourism and culinary experiences.