News

The KDD-2001: ACM 7th annual international conference on Knowledge Discovery and Data Mining Conference was held in San Francisco, CA, Aug 26-29, and was attended by over 800 participants from 20+ countries.

Among the highlights of the conference were:

* An excellent technical program, featuring the best 20 papers and 32 posters chosen from over 200 submissions.

* 6 tutorials on E-Business Enterprise Data Mining, Scalable Frequent-Pattern Mining Methods, Data Mining for Outliers, Value-based Data Mining for CRM, KDD for Mobile and Distributed Environments, Advances in Decision Tree Construction

* 6 full-attendance workshops

* 3 very lively Panels on When and How to Subsample, Data Mining Startups: The Perfect Storm and New Research Directions in KDD

* keynotes talks by Raghu Ramakrishnan, CTO and Founder, QUIQ, about Mass Collaboration

Tom Mitchell, WhizBang! Labs and CMU, about extracting useful information automatically from web pages using co-training. FlipDog job site created using this method was recently bought by Monster.com -- a visible success of machine learning.

Russ Altman, Professor of Medicine and CS at Stanford, energized the audience with his presentation on Challenges for Knowledge Discovery in Biology.

KDD-Cup 2001 drew over 130 groups (estimated 300-400 people) who participated to produce a total of 200 submitted predictions over the 3 tasks in bioinformatics -- see http://www.cs.wisc.edu/~dpage/kddcup2001/ for full results.

Excellent research and industry track papers were complemented by 29 exhibits.

Bioinformatics and text mining were among the emerging directions for data mining.

Look for further reports from KDD-2001 conference, including panel summaries in future issues of KDnuggets News.