A Picture-based Approach to Travel Recommender Systems
Personalized recommendations strongly rely on an accurate model to capture user preferences. Eliciting this information is, in general, a hard problem. In the field of tourism, this initial profiling becomes even more challenging. It has been shown that particularly in the beginning of the travel decision making process, users themselves are often not conscious of their needs and are not able to express them. Aiming at revealing implicitly given user preferences, this work introduces an approach that utilizes a set of travel related pictures to discover users’ travel behavior and in turn, to deliver recommendations. Next, a more general approach based on convolutional neural networks is introduced, where any set of pictures can be used to characterize both travelers and tourism destinations. This talk discusses a stream of studies to quantify intangible user preferences and to provide easy and playful methods to generate inputs/data for recommendation systems.