Group recommender systems for tourism: current state and future directions

Abstract

Recommender systems, as sucessful personalization tools, are focused on helping potential clients to obtain the information that best corresponds to their needs and preferences, in a search space overloaded with options. Properly, group recommender systems (GRSs) focus on suggesting certain types of item (products or services) that tend to be consumed in groups. This paper analyzes the use of GRSs in the tourism domain. It considers the most representative works in this area, highlighting the type of study and evaluation, the application scenario, its purpose, and the methodology used for the generation of recommendations. With this goal, the PRISMA methodology is used, being 1115 works initially identified in the Web of Science and Scopus, which are reduced into a set of 79 articles after the application of several inclusion and exclusion criteria. A comprehensive study of such works identifies three main working domains: 1) GRSs for restaurants, 2) GRSs for itineraries and points of interest (POIs), and 3) GRSs for destination and travel. An overview of the main methodologies and evaluation approaches in these works is also discussed. The analysis performed also identifies relevant research lines for future work.