A Personalized System for Tourist Attraction Recommendation

Abstract

Currently, travelers can easily access travel information from the internet by themselves.Unfortunately, most of these tourism websites recommend the same content to every traveler.Hence, travelers may receive overwhelming number of options or may receive information that does not comply with their own interests.This article proposes a personalized system for tourist attraction recommendation using a clustering technique and an analytic hierarchy process (AHP).The clustering and AHP are combined to construct a ranking model of tourist attractions.The model is used to recommend the tourist attractions based on individual traveler's preferences and constraints.The ranking model evaluation uses 400 test cases consisting of 50 tourist attractions ranked by 400 travelers.The evaluation results reveal that the tourist attraction ranks obtained from the ranking model is similar to those ranks rated by experienced travelers.The Spearman correlation coefficient is equal to 0.907.