Discovering implicit activity preferences in travel itineraries by topic modeling

Travel itineraries are employed in tourism research to study tourist activities for various applications. However, the potentials of such itineraries in providing insights into the activity preferences of tourists have not been explored because of the complexity of travel information. In this paper, a new approach based on probabilistic topic modeling with latent Dirichlet allocation is introduced for travel itinerary analysis and representation. Capable of revealing the implicit preferences of tourists, the new approach enables topic modeling to be applied in itinerary analysis. We demonstrate its effectiveness through a case study of outbound travel behavior analysis on a large-scale travel itinerary data set. Activity profiles of various itinerary types at different destinations are revealed. The results are useful for travel and tourism managers in developing travel and tour packages. The general features of the proposed method can be applied into different tourism contexts and travel itinerary formats for wide applications.