Recommender systems, cultural heritage applications, and the way forward.

Abstract Modern recommender systems appeared during the last decade of the 20th century and have already proven their importance in tackling with information overload and content or services personalisation in the modern big-data world. This is markedly suggested by their adoption by major industry parties in a wide variety of domains. The application of recommender systems in the Cultural Heritage domain has appeared relatively recently, particularly during the advent of the 21st century. Nevertheless, a large number of works appeared in the scientific literature, driven primarily from large-scale cultural heritage collaborative projects. This paper presents a review of the current recommender system approaches, including some basic mathematical formalism, and focuses on recommender system applications in Cultural Heritage, to identify the advantages, the benefits and prospects, the disadvantages and pitfalls. The paper discusses the diversity and some common assumptions in the reviewed works, outlines some proposals for future development, and takes a position on the way forward. [ABSTRACT FROM AUTHOR]