Features

Recently we saw several major scientific and engineering advances that were stimulated by a competition to solve grand challenge problems. The DARPA Grand Challenge produced great advances in robotic car navigation in 2005. Ansari X-prize was awarded in 2004 for the first successful commercial spaceflight. RoboCup, whose goal is to develop a team of humanoid robots that can win against the human world soccer champion team by 2050, has greatly advanced robotic performance and created many robotic enthusiasts.

Is there a Grand Challenge Problem for Data Mining?

A Panel at KDD-2006 Conference Panel will address this question.

If you have ideas for Grand Challenge that fit criteria below, please email to gregory at kdnuggets.com and your ideas will be considered for panel discussion, a planned follow-up article, and, perhaps could lead to a great advance in data mining !

A good grand challenge problem for data mining should satisfy several criteria:

• Data mining should be central for the solution
• Public: the data set should be public
• Specific: there should be specific goal (preferably stated in one sentence)
• Hard: The problem should be hard to solve
• Interesting: the problem should be interesting and exciting to attract researchers, press, public, and funding

• Automatic tagging and classification of digital photos and images on the web
• Identifying all genes and potential therapeutic targets for cancer
• A text-mining and understanding system that can use the web to pass standard tests like SAT
• Ranking Wikipedia pages for trust/authority