Package recommender systems: A systematic review

Issue title: Special issue: Surveys in Artificial Intelligence-based Technologies

Article type: Research Article

Authors: van Schaik, S.N.^{a; *} | Masthoff, J.^a | Wibowo, A.T.^b

Affiliations: [a] Utrecht University, Utrecht, The Netherlands | [b] Institut Teknologi Telkom, Bandung, Indonesia

Correspondence: [*] Corresponding author: S.N. van Schaik, Utrecht University, Utrecht, The Netherlands. E-mail: [email protected].

Abstract: This paper presents a systematic review of research into package recommender systems (PRS). PRS recommend a combination of items rather than individual items, for example, multiple attractions to visit on a day trip to a city, trousers and a shirt to wear together, or multiple dishes to make up a meal. The review provides a framework for considering existing PRS research, and highlights the techniques used for package recommendation and the evaluation methods and metrics employed. It also raises many issues that warrant future research.

Keywords: Bundle recommendation, package recommendation, constraints, recommender system, Bandung

DOI: 10.3233/IDT-190140

Journal: Intelligent Decision Technologies, vol. 13, no. 4, pp. 435-452, 2019