A Model-Driven Approach to Evolve Recommender Systems

Resumo

Recommender systems have become an important issue on Web applications, but its research is usually focused on algorithms and data optimization. However, as the recommendation techniques improve and these systems become more commonly used in software applications, there is the need of easily adapt and evolve them. To address this need, we propose a model-driven approach to evolve recommender systems and present an architecture solution from our research, using an events management system as the domain for an use-case scenario. Future work might demonstrate the architecture feasibility.