New digital technologies support personalized recommender systems that can assist a tourist who wants to make a city tour. To develop a smart system that can give tourists an optimized complete activity program for their trip, it is not only important to know the preferences and interests of tourists but also whether they like combinations of activities/points of interest (POIs) or not. The aim of this study is to measure and predict tourists’ preferences for combinations of activities in planning a program during a city trip. A personalized stated choice experiment is developed and presented in a survey to a random sample of 238 respondents. Binary mixed logit models are estimated on the choice data collected. An advantage of this approach is that it allows estimation of covariances between city trip activities indicating whether they would act as complements or substitutes for a specific tourist in his/her city trip activity program. The model parameters provide information on combinations of activities and themes that tourists prefer during their city trip and that the recommender system can use to further fine-tune the recommendations of city trip programs and optimize the tourist experience.
|Title of host publication||Trends in Tourist Behaviour: New Products and Experiences in Europe|
|Editors||Andres Artal-Tur, Nazmi Kozak, Metin Kozak|
|Place of Publication||Cham|
|Publication status||Published - 2019|