Mas Nurul Hamidah, Triyanna Widiyaningtyas, Wahyu Sakti Gunawan Irianto, Rahmawati Febrifyaning Tias, Salahudin Robo
Digital platforms in the tourism sector are expanding rapidly, leading to information overload when travellers attempt to select destinations that match their preferences. Recommendation systems based on Collaborative Filtering (CF) have emerged as an effective solution to this challenge. However, the optimal similarity measure for sparse, user-generated tourism data remains unclear. This study compares the performance of two similarity algorithms, Jaccard and Hamming, within a user-based Collaborative Filtering framework using a public TripAdvisor dataset representing tourist preferences. The evaluation metrics employed include Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Precision. The experimental results show that Hamming similarity outperforms Jaccard, demonstrating greater robustness across datasets of varying sizes. Specifically, the Jaccard algorithm achieved an MAE of 0.7906 and RMSE of 1.0868 on the 10K dataset, while the Hamming algorithm maintained lower error values and higher precision across both datasets. These findings confirm that Hamming's bitwise comparison mechanism provides more stable and accurate predictions for sparse tourism data. © 2025 IEEE.
Malang State University, Department of Electronics and Informatics Engineering, Malang, Indonesia; University of Bhayangkara, Informatic Engineering, Surabaya, Indonesia; Universitas Yapis Papua, Department of Information Systems, Indonesia
