Artificial Intelligence Data Science


Recommendations can be classified using so-called recommendation logics. There are five different recommendation logics:

  • Product-related: Recommendations that relate to one or more products for which the customer is interested.
  • Category related: Recommendations that relate to the category that the customer is interested in.
  • User-related: Recommendations that are based on the current buying behavior of the user as well as on “historical” data of the user, eg. as the purchase history refer.
  • Keyword-related: Recommendations that refer to the current search term of the user with whom he triggered the search (eg in the online shop).
  • Global: Recommendations that relate to the entire range.

The recommendation logic describes the reference system of a recommendation. You often see the different types of recommendations in certain places in online shops. “Other Customers Also Bought” usually shows product-related recommendations, while the headline “your personal recommendations” refers to user-related recommendations.