On the End-to-End Development of a Cultural Tourism Recommender

Abstract

Recommenders are systems that employ some knowledge on items and user preferences, along with sophisticated algorithms to provide personalised content and services. They have been around to tackle the information overload and personalisation demand in today's always-connected world. This technology appeared in the cultural heritage domain relatively recently, but the bibliography is already rich, as cultural tourism plays an important role for regional economies. From the technical perspective, different approaches, like collaborative filtering, content-based, knowledge-based and hybrid approaches, have been adopted. From the intuition perspective, the approaches are influenced by current conceptualisation and specific application domains and demands. The museum has been one of the main target applications, either as a part of visit support or in the context of cultural tourism initiatives. This article presents a review of the domain and draws a generic blueprint for the end-to-end development of a recommender for cultural tourism that outperforms a baseline popularity-based approach.