News

In the context of the recent posting on the role of statistics
in data mining, readers of this list may be interested in reading
two recent review papers that discuss some of the positive
aspects of the relationship between modern data mining and statistics:

1. "Data mining at the interface of computer science and statistics,"
2001, revised version in press. This paper is primarily intended
for a computer science and engineering audience.

2. "Data mining: data analysis on a grand scale?"
Technical Report UCI-ICS 00-20, July 2000. A revised
version of this appeared in the journal Statistical Methods in Medical
Research, and is written for a statistical audience explaining
how data mining is different from traditional statistics.

Both papers are online at:
http://www.datalab.uci.edu/papers.html#massive-data

The bottom line is that statistics has much to offer computer
science and computer science has much to offer statistics.
Perhaps the greatest need is in education, namely the need to
educate more students at the intersection of both disciplines.

Padhraic Smyth
Associate Professor
Information and Computer Science
University of California, Irvine
CA 92697-3425
smyth@ics.uci.edu