PROGRAMME

The Programme for PMKD4 '06 is yet to be constructed, based on papers submitted and accepted following peer review. PMKD4 '06 will be a one or two day Workshop consisting of 10 or 20 accepted papers respectively. Exact dates are not yet known.

In this, the first International PMKD DEXA Workshop we have papers from the leading
experts from around the globe. Unfortunately, several of our colleagues who would have
liked to have been with us could not be so for unavoidable reasons.

We face complex diagnostic and prescriptive problems. The relatively new discipline
of Data-Mining (DM) and Knowledge Discovery in Databases (KDD) is clearly ailing.
The discipline has claimed much for itself in Business Applications and Business Intelligence,
but seems to be delivering only partial solutions. Research activity seems to be entirely
technical and it there seems to be no rationale, and no direction. Finally, DM & KDD,
being formed from the confluence of the disciplines of Statistics, Machine-Learning and
Database is a complex ‘organism’ with many internal contradictions and clashes of paradigm.

One of our seemingly most simple problems lies in the use of the same terms for similar
but often subtly different concepts within the constituent disciplines. We know the solution to
this problem, don’t we? “Doctor, Heal Yourself!” This is the classic problem of metadata/
nomenclature, which itself has a central part in each of the constituent disciplines. However,
this edict is much easier to state than to implement. Disciplines are like colonial empires.
They grow by assimilating the concepts and techniques of their sister disciplines, and
this is especially so with our three foundation disciplines. Hence the ownership of terms,
concepts and methodologies becomes unclear and each of us, as loyal subjects of
our own empires, believe that real truth and reason lies within our own empire. We know
we need a common metadata language, but who should define it, and who should arbitrate
on disputes? It seems that a meta-metadata-language is the only way forward. A meta
metadata language, or ontology, which encompasses all, and which is greater than all.
But who amongst us is capable of learning and speaking such a language?
If DM&KDD is really to become a coherent discipline, then we all need to!

However, the terminological clashes are not the most problematic. Differences in concepts
and paradigms for looking at exactly the same issue mean that an integrated DM&KDD disciple
must be multi-paradigmic if it is to survive. We must therefore take care that we do not
fall into the trap of thinking that a single paradigm-methodology-philosophy will suffice for our
amalgamated discipline.

Prescribing a solution for the ails of DM&KDD therefore poses major philosophical and
social challenges to us, the members of our new amalgamated empire. We need to be
aware where we each come from, personally and professionally, when we propose what
we think are the right philosophies and methodologies for DM&KDD. We need to be aware
that all proposals are likely to be partial perspectives, and to accept that we need to operate
in a liberal philosophical and methodological space in which the clashes of philosophies
produce new syntheses rather than retrogressive fragmentation.

These philosophical and social challenges face us in PMKD. Let’s face them together as
a family, and let our family grow!

10.40-11.20     Keith Rennolls, University of Greenwich, U.K.

OPENING        An Intelligent Framework (O-SS-E) for Data Mining, Knowledge Discovery

                          and Business Intelligence

11.20-11.55    Sylvain Delisle, Université du Québec à Trois-Rivières, Canada

                          Integrating Data Mining and Decision Support via Computational Intelligence: 
 Towards Significant Progress in Applied Data Mining

11.55-12.30 Evgenii Vityaev & Boris Kovalerchuk (Sobolev Institute of Mathematics,
Russia & Central Washington University, USA)
Relational Methodology for Data Mining and Knowledge Discovery

14.00-14.40     Mykola Pechenizkiy¹, Seppo Puuronen¹ & Alexey Tsymbal² (¹University of Jyväskylä,
                          Finland & ²Trinity College Dublin, Ireland)
INVITED       Competitive advantage from Data Mining:
                         Lessons learnt in the Information Systems field.

14.40-15.15    John Strassner & B. J. Menich, Motorola Labs, USA.
                        Philosophy and Methodology for Knowledge Discovery
                         in Autonomic Computing Systems

15.15-15.50 Alun R. Butler, University of Greenwich, U.K.
Three Dogmas of Metadata and Undiscovered Knowledge

Open discussion: The way forward for PMKD?
Construction of an agenda for the future.

Philosophies and Methodologies for Knowledge Discovery, Deployment, and Development of Decision Support Systems (PMKD4'06)