The Jester data set is a set of user ratings of jokes collected from
the Jester joke recommender system [50]. It consists of two data sets:
one has ratings of 100 jokes from 73,421 users between April 1999 and
May 2003, and the other has ratings of 150 jokes from 63,974 users
between Nov 2006 and May 2009. These ratings are continuous ratings
in the range [−10,10].3 As a result of the “gauge set” required by the
Eigentaste algorithm, a portion of the Jester data set is dense: there is
a small set of jokes that almost every user has rated.