|
When in 1986 Yves Kodratoff started the European Working Session on Learning
at Orsay, France, it could not be foreseen that the conference would grow
year by year and become the premier European conference of the field, attracting
submissions from all over the world. The first European Conference on
Principles of Data Mining and Knowledge Discovery was organized by Henryk
Jan Komorowski and Jan Zytkow in 1997 in Trondheim, Norway. Since 2001 the
two conferences have been collocated, offering participants from both areas the
opportunity to listen to each other’s talks. This year, the integration has moved
even further. Instead of first splitting the field according to ECML or PKDD
topics, we flattened the structure of the field to a single set of topics. For each of
the topics, experts were invited to the Program Committee. Submitted papers
were gathered into one collection and characterized according to their topics.
The reviewers were then asked to bid on all papers, regardless of the conference.
This allowed us to allocate papers more precisely.
The hierarchical reviewing process as introduced in 2005 was continued. We
nominated 30 Area Chairs, each supervising the reviews and discussions of about
17 papers. In addition, 307 reviewers completed the Program Committee. Many
thanks to all of them! It was a considerable effort for the reviewers to carefully
review the papers, some providing us with additional reviews even at short notice.
Based on their reviews and internal discussions, which were concluded by
the recommendations of the Area Chairs, we could manage the final selection for
the program. We received 521 submissions, of which 100 were presented at the
conferences, giving us an acceptance rate of 20%. This high selectivity means,
on the one hand, that some good papers could not make it into the conference
program. On the other hand, it supports the traditionally high standards of the
joint conference. We thank the authors from all over the world for submitting
their contributions!
Following the tradition, the first and the last day of the joint conference were
dedicated to workshops and tutorials. ECML PKDD 2008 offered 8 tutorials and
11 workshops.We thank theWorkshop and Tutorial Chairs Siegfried Nijssen and
Arno Siebes for their excellent selection. The discovery challenge is also a tradition
of ECML PKDD that we continued. We are grateful to Andreas Hotho and
his colleagues from the Bibsonomy project for organizing the discovery challenge
of this year. The results were presented at the Web 2.0 Mining Workshop.
One of the pleasures of chairing a conference is the opportunity to invite
colleagues whose work we esteem highly. We are grateful to Fran¸coise Fogelman
Souli´e (KXEN) for opening the industrial track, Yoav Freund (University of
California, San Diego), Anil K. Jain (Michigan State University), Ray Mooney
(University of Texas at Austin), and Raghu Ramakrishnan (Yahoo! Research)
for accepting our invitation to present recent work at the conference. |
