While there, follow the link to the Data Mining Presentations where
you will find information about the KDD-Cup-97, the KDD-Cup-98 and
the KDD-98 Exhibits Presentations.
Among other things, the KDD-Cup-98 page now includes the data set
(number of records: 190,000+, number of variables: 481), the
documentation released for the competition, and the KDD-Cup
presentation I made during the KDD-98 conference.
The KDD-Cup data set and the accompanying documentation are available
for general use with the following restrictions:
(1) The users of the data must notify
Ismail Parsa (iparsa@epsilon.com) and
Ken Howes (khowes@epsilon.com)
in the event they produce results, visuals or tables, etc. from
the data and send a note that includes a summary of the final
result.
(2) The authors of published and/or unpublished articles that use
the KDD-Cup-98 data set must also notify the individuals listed
above and send a copy of their published and/or unpublished work.
(3) If you intend to use this data set for training or educational
purposes, you must not reveal the name of the sponsor PVA
(Paralyzed Veterans of America) to the trainees or students. You
are allowed to say "a national veterans organization"...
In the case you are not able to email the results or the articles,
please mail or fax them to the following address:
Ismail Parsa
Epsilon
50 Cambridge Street
Burlington MA 01803 USA
TEL: (781) 685-6734
FAX: (781) 685-0806
P.S. The target fields that were left out of the validation data
set for the purposes of the competition are now in the file
'VALTargt.txt'. These fields were used by the KDD-CUP-98
Committee to evaluate the accuracy of the predictions
submitted by the participants.
Yours,
Ismail