Development Tourism Destination Recommendation Systems using Collaborative and Content-Based Filtering Optimized with Neural Networks

Abstract

Tourism, a vital sector in the global economy, benefits significantly from advancements in infrastructure, accessibility, and information availability. However, the vast volume of information can overwhelm travelers, underscoring the need for efficient recommendation systems. This research aims to develop an advanced tourist destination recommendation system by integrating Collaborative Filtering (CF) and Content-Based Filtering (CBF) models with Neural Networks. This approach seeks to enhance recommendation accuracy by closely aligning with user preferences and addressing the challenge of limited data. The study utilizes data from 523 tourist destinations across West Java, along with user preference assessments, encompassing stages of data collection, labeling, pre-processing, pre-training, neural network-based training, model evaluation, and the presentation of recommendation outcomes. The optimization of the recommendation models through neural networks has notably improved the precision of tourist destination suggestions, achieving Root Mean Square Error (RMSE) values below 0.37 for both CF and CBF approaches. This research significantly contributes to increasing the search efficiency and accuracy for tourist destination information, offering a strategic solution to the prevalent issue of information overload in the tourism industry.