Spring 2006 DLF Forum: Melvyl Recommender Project,
34
3 April, 2006
Finding Relationships
Patrons
Related Items
Started to experiment
Item => checked out by four different patrons.
Follow links =>
set of related items.
How to narrow to usable recommendations?
Sets often hundreds of items.
User studies:
3 -5 recommendations are optimal.
How do we prioritize?
