CFP

workshop "Data Mining for Marketing Applications"
will be held at the ECML/PKDD 2001 this september in Freiburg/Germany.

Scope of the Workshop:

Contributions of data mining and statistics with a clear reference to
marketing problems describing practical experiences and requirements,
theoretical work in progress as well as position papers are invited.

All contributions must focus on topics that are of special importance
for marketing applications and that require special measures. We also
encourage papers showing marketing problems not yet resolved
sufficiently and approaches successfully deployed in other disciplines
that can be transferred to marketing.

Of particular interest are methods and proposals that address the following
issues:

1.  Transformation of marketing problems (e.g. the acquisition of new
customers) into data mining goals (e.g. finding differences between
profiles of customer and non-customers via a classification or
identifying groups of homogeneous customers via segmentation).

2.  Integration of internal and external data sources to generate an
appropriate target and contrast group regarding legal, technical and
organizational constraints.

3.  Evaluation of alternative modeling strategies, assessment of model
quality and identification of measures to select a "best" model
regarding the business goal.

4.  Process learning from one marketing campaign to another: Creation
of a proper test design, response analysis, problem of statistical
validity when generating models in several loops, evaluation of a data
mining project in order to install a continuous learning cycle.

5.  Integration of domain dependent marketing expertise into data
mining techniques to increase utility of identified knowledge:
e.g. integration of (subjective) retail background knowledge into the
data mining process in order to increase the usability of extracted
association patterns for marketing decision making.

6.  Web mining techniques to support market-oriented and personalized
communication with website visitors: e.g. clustering of website
visitors based on their clicking behaviour, or categorizing visitors
according to their position in the customer lifecycle
(i.e. acquisition, fostering, retention) in order to enable
personalized targeting.

Papers related to this subject must be send until the June, 8th.

More information can be found at:
http://www.riskgroup.de/pkdd/workshop.htm and
http://www.luc.ac.be/iteo/events_workshop.html