Tourism Recommendation System using User Based Collaborative Filtering
Kurniawan Eka Permana, Abdullah Basuki Rahmat, Eka Mala Sari Rochman, Aery Rachmad, Sigit Susanto Putro

Department of Informatic, Faculty of Engineering, University of Trunojoyo Madura

Abstract

This study investigates the design and evaluation of a tourism recommendation system based on user-based collaborative filtering. The study makes use of the Indonesia tourism destination dataset obtained from Kaggle, which includes user ratings for a diverse range of tourist destinations around Indonesia. The dataset, which includes ratings from 300 users on a scale of 1 to 5, serves as the foundation for the proposed methodology. The approach is made up of several key stages, beginning with data loading and transformation into a matrix format. To improve suggestion accuracy, user ratings are normalized based on the average rating of each user, taking into consideration individual rating tendencies. Then, cosine similarity measurements are used to identify people with similar tastes, with missing values addressed before calculating similarity scores. The method suggests tourism destinations to users during the recommendation phase using a weighted average of user similarity scores and place ratings. The performance of the recommendation system is evaluated by calculating Root Mean Square Error (RMSE) and analyzing prediction accuracy. Several scenarios are thoroughly investigated, each with a different number of top neighbors considered for predictions. Surprisingly, the scenario containing the top 20 neighbors produces the best results, with the lowest RMSE of 0.3646, showing much improved prediction accuracy. The methodology has potential for additional developments in tourism recommendation systems, with the RMSE metric confirming the system^s effectiveness

Keywords: Recommendation System- Collaborative Filtering - User Based Collaborative Filtering

Topic: Artificial Intelligence and Data Science