KDD Cup 2012 Call for Proposals

We invite proposals for KDD Cup 2012 competition. KDD Cup is the well-known data mining competition of the annual ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD). The KDD-2012 conference will be held in Beijing, China, August 12-16 2012. One proposal will be selected for the 2012 competition, and one member of the organizing team will be invited to be the second KDD Cup 2012 co-chair.

Proposals should include a short paragraph on each of the following items:

  1. Description of the problem addressed, with general background information on the application domain.
  2. Description of the available data, guarantee of availability, guarantee of confidentiality of the "ground truth", and size.
  3. Description of the competition tasks, their scientific and technical merit and their practical significance.
  4. Description of the evaluation procedures and established baselines (provide evidence that non-trivial performance can be achieved, and an estimate of what constitutes a significant difference in performance). The evaluation metrics should be both meaningful for the application and statistically sound for objective performance comparison.
  5. List of the available resources (team member availability, computers, support staff, other equipment, sponsors). Please state whether you might need help with the infrastructure of hosting the competition including a site to sign-up teams, accept submissions, host a leaderboard, etc.
  6. Schedule. The competition should last between 6 and 8 weeks and the winners should be notified by mid-June. The winners will be announced in the KDD-2012 conference.
  7. Names, affiliations, postal addresses, phone numbers, and short biographies of the organizers.
  8. Whether the competition is new or has been held before.
  9. Whether you will award prizes of any form to winning teams (e.g., sponsor the conference registration, monetary reward).

A good competition task is one that is practically useful, scientifically or technically challenging, can be done without extensive application domain knowledge, and can be evaluated objectively. Of particular interests are non-traditional tasks/data that may need novel techniques and solutions. Please send your proposals to briand@gmail.com by Nov 15, 2011.