Mediation of User Models: for Enhanced Personalization in Recommender Systems

Provision of personalized recommendations to users requires accurate modeling of their interests and needs. This book presents a general framework and specific methodologies for enhancing the accuracy of user modeling in recommender systems by importing and integrating data collected by other recommender systems. Such a process is defined as the mediation of user models and user modeling data. This book discusses the details of the generic user modeling mediation framework, provides a user modeling data representation model, demonstrates the compatibility of the model with existing recommendation techniques, and discusses the general steps of the mediation. Then, the mediation framework is applied and illustrated with two practical mediation scenarios: cross-technique mediation and cross-domain mediation. Empirical evaluation of these scenarios shows that the mediation of user modeling data is practical and beneficial, as it allows upgrading the quality of the recommendations provided to the users.