Knowledge-based recommenders use background knowledge about associated
and similar items to infer the needs of the user and how they can best be met.
Knowledge-based methods will then draw not only on typical measures of simi-
larity like correlation, but also on feature similarities that will interest the user. For
instance, when a user indicates that he liked A Good Year, a KB recommender sys-
tem might know that this film could be associated with either A Beautiful Mind
(which also stars Russell Crowe) or Jeux d’Enfants (which also stars Marion Cotil-
lard). Since the user has shown a preference for French films in the past, the system
will assume that the user liked A Good Year because it featured Marion Cotillard,
and recommend accordingly. It is pointed out in [16] that KB recommenders often
draw on case-based reasoning approaches.