Since pure recom- mendation systems contain multiple weaknesses that can usually be tackled by merging different paradigms, hybrid systems have become the current most popular choice, especially when the system needs to deal with highly heterogeneous information.