Topics: AI | Data Science | Data Visualization | Deep Learning | Machine Learning | NLP | Python | R | Statistics

KDnuggets Home » News » 2019 » Feb » News » A Word from KDD Cup 2019 Organizers

A Word from KDD Cup 2019 Organizers

This year’s KDD Cup will be celebrating 22 years. It’s been an exciting journey and we have come a long way! We invite industrial and academic institutions to submit proposals for organizing the 2019 KDD Cup Competition. Learn more now!

By KDD Cup 2019 Organizers.


Challenging a community of experts and enthusiasts to solve a scientific problem has been around for hundreds of years. One of the first recorded challenges like this was established in 1714 by British Board of Longitude. The prize was awarded to the person who could solve what was arguably the most important technological problem of the time: to determine the longitude of a ship at sea.[1] This prize was motivated after a disaster in 1707 at Scilly where 4 ships were drowned. The Board administered prizes for those who could demonstrate a working device or method. The max prize for a method that could determine longitude within 30 nautical miles (56 km; 35 mi) (£2,600,000 as of 2015) [2].

KDD (Knowledge, Discovery and Data Mining), an ACM Special Interest Group, has pioneered the data science competition, known as “the KDD Cup”. This year’s KDD Cup will be celebrating 22 years. It’s been an exciting journey and we have come a long way!

Many accomplished leaders in the field have contributed in organizing this event over the years. They shaped it to reflect the most interesting domains and problems, existing or emerging; uncover best methodologies and tools, attract and discover great minds, boost teamwork and creativity in contenders, let them leverage their problem-solving skills, and, more importantly, bring up new frontiers.

In these last two decades, KDD Cup have covered:

  • forecasting of air quality indices (2018),
  • highway tollgates traffic flow prediction (2017),
  • measuring impact of research institutions via prediction of scientific research papers acceptance rates and citations (2016),
  • prediction of course drop-outs for college students (2015),
  • prediction of fulfilled requests on educational materials made by school teachers to improve funding outcomes (2014),
  • resolving author-name ambiguity in scientific publications (2013),
  • prediction of followers and click-through rate to improve user engagement with the online content in social networks (2012),
  • identification of user tastes in music for Yahoo! music recommendation (2011),
  • prediction of student’s performance for improvements in education quality (2010),
  • prediction of customer relationship for personalization in CRM (2009),
  • early breast cancer detection from medical imaging (2008),
  • user rating prediction to improve Netflix movie recommendation (2007),
  • pulmonary embolism detection from medical imaging data (2006),
  • categorization of search queries from an internet search engine (2005),
  • plus protein homology prediction in particle physics (2004),
  • prediction of paper citations and building citations graph for arXive (2003),
  • detect experimental evidence of gene expression from microbiology research papers for Flybase using their gene-expression curation criteria (2002),
  • prediction of molecular bioactivity and plus protein locale for drug design (2001),
  • clickstream analysis for online retailer web-site (2000),
  • prediction of attack type in network intrusion detection (1999),
  • identify response to the mailing in direct marketing for profit optimization (1998, 1997)

We invite industrial and academic institutions to submit proposals for organizing the 2019 KDD Cup Competition. We are looking for complex data science problems with high novelty, societal impact, and/or economic value. Interested organizers please check the Call for Proposal page, as well as a more detailed formatting guideline via the KDD Cup 2019 website.

This year is super exciting for us, as it is the first time we will have an auto-ml competition in addition to a regular competition. Organizers interested in knowing more about the auto-ml competition should check the “Auto-ml FAQ” Appendix to the Call for Proposals. Additionally, we are in conversations with a few potential platform sponsors, who may be able to host these competitions on their websites. Therefore, if you have novel problems and data at hand without platform or hardware support, there will be some options available to you.

To discuss ideas or to ask specific questions, please contact us at: [email protected]. Please note that proposals are being accepted through February 17th, 2019.


By subscribing you accept KDnuggets Privacy Policy