The stereotype process and the hybrid recommender algorithm are unique features of the YOOCHOOSE technology. They facilitate the use of simple content- or user-based recommendation algorithms where appropriate and to skillfully mix them with each other and the processes of the YOOCHOOSE platform via the hybrid approach.
Basically, recommendation algorithms can be categorized as either content-based or collaborative algorithms. In each of these categories, the YOOCHOOSE system uses several implementations to compute the best recommendations.
Depending on their application and the current context, content-based and collaborative as well as stereotype algorithms have both advantages and disadvantages. Based on the hybrid process used by YOOCHOOSE, these can be cleverly exploited, thereby enhancing the quality of recommendations.
The stereotype process uses the existing knowledge about a domain, which is available in most applications, in order to classify the users or items into meaningful clusters, thereby attaining a new quality in recommendations.
To be able to offer its customers the best recommendation processes at all times, YOOCHOOSE is continually evolving its recommender system. The next expansion will facilitate recommendations that are optimally matched to the user's context.

