Bootstrapping recommender systems based on a multi-criteria decision making approach

Recommender Systems (RSs) cope with the problem of information overload, by providing to users content that fit with what they prefer. Generally, RSs work much better for those users on which they have more information about. Satisfying the new users becomes a challenge, as ensuring for them recommendations of quality is vital for the growth of the RS. Coping with this issue can be made by ensuring a certain brief interview with the user-called bootstrapping process-through which we acquire a user's feedback on a set of items, to subsequently enrich the user profile and inferring efficient recommendations. In this paper, we will propose an approach for bootstrapping a RS based on multi-criteria ratings and a method of computing weights of criteria taken from Multi-criteria Decision Making (MCDM).