Tourism destination Recommender Systems: A Scoping Review and Thematic Clustering of Scientific Literature

Tourism destination recommender systems (TDRSs) have become pivotal for reducing information overload and supporting personalized destination choice in digital tourism ecosystems. However, the field remains methodologically fragmented, spanning heterogeneous data sources (e.g., reviews, clicks, contextual signals), modeling paradigms (e.g., hybrid, semantic/ontology-based, deep learning), and evaluation practices. This scoping review maps the research landscape and derives an evidence-guided thematic structure of TDRSs. We conducted a structured search in the Web of Science Core Collection using the query “tourism destination” AND “recommender system,” yielding a corpus of 223 records. Following PRISMA-ScR principles, we charted bibliographic metadata, publication types, venues, and author keywords/Keywords Plus. To generate an interpretable thematic map, we applied (i) co-word analysis on combined keyword fields to detect topical communities, (ii) bibliographic coupling to identify research fronts, and (iii) a narrative synthesis of representative papers per cluster. The resulting taxonomy highlights recurring clusters such as context-aware/mobile destination guidance, semantic and ontology-driven recommendation, hybrid and multi-criteria decision support, social media and sentiment/emotion-aware recommendation, and deep learning for intent-aware travel recommendation. By consolidating these clusters, the review clarifies how methodological choices align with tourism decision constraints (time, location, weather, group travel) and provides a roadmap for future evidence-based studies, including reproducible benchmarks, explainability, and robust evaluation protocols.