Solving Package Recommendation Problems with Item Relations and Variable Size

CHRISTIAN VILLAVICENCIO; SILVIA N. SCHIAFFINO; JORGE ANDRES DIAZ PACE

Rosario

Simposio; Argentine Symposium on Artificial Intelligence (ASAI 2015) - Parte de Jornadas Argentinas de Informática e Investigación Operativa (JAIIO); 2015

SADIO

In this article, we explore an approach to solve the problem of recommending a package of items (each of them with a score and a cost) to a user.In our approach, we consider two types of relations between items: dependency and incompatibility; and we also consider that the size of the package is not fixed but cost-driven. To this end, adaptations of existing package recommendation algorithms are proposed. We have evaluated the proposed approach in a specific domain and obtained promising results.