Use and number of archetypes | HCL Digital Experience
To generate recommendations for items that no mentors have rated, the Preference Engine uses synthetic users called "archetypes." These users possess a particular characteristic very strongly.
For example, an archetype who only likes action movies might rate all action movies very highly. Recommendations from archetypes are added to recommendations produced by mentors.