Optimizing Personalized Ranking in Recommender Systems with Metadata Awareness
Marcelo Garcia Manzato, Marcos Aurélio Domingues, Solange Oliveira Rezende · 2014 IEEE/WIC/ACM International Joint Conferences on Web Intelligence (WI) and Intelligent Agent Technologies (IAT) · 2014
In this paper, we propose an item recommendation algorithm based on latent factors which uses implicit feedback from users to optimize the ranking of items according to individual preferences. The novelty of the algorithm is the integration of content metadata to improve the quality of recommendations. Such descriptions are an important source to construct a personalized set of items which are meaningfully related to the user's main interests. The method is evaluated on two different datasets, being compared against another approach reported in the literature. The results demonstrate the effectiveness of supporting personalized ranking with metadata awareness.