Leave a comment

The 2nd International Workshop on the Induction of Process Models (IPM'08)
at ECML PKDD 2008, 15 September 2008, Antwerp, Belgium

Call for Papers

While the worlds of science and business typically meet in the presence of
a profitable scheme, individuals from both environments have interests in
analyzing complex data about dynamic systems. Whether motivated by a drive
to increase system efficiency or to understand nature, their shared goal
leads to a shared focus on the underlying causal processes that explain or
produce observed phenomena. To this end, researchers construct models from
data derived from observed system behavior and background knowledge about
the candidate processes. Traditional literature on regression, time-series
analysis, and data mining produces descriptive models that may reproduce
the observed data but cannot explain the principal dynamics. Therefore,
researchers are called to develop methods that capture complex temporal
and spatial relationships in terms of domain knowledge (e.g., relevant
scientific or business concepts) and that construct these explanatory
process models.

One can develop both qualitative and quantitative process models depending
on their intended use. Qualitative approaches to model induction include
learning state transition models, Petri-nets, and learning from
(time-stamped) event sequences and event logs. Qualitative representations
are particularly interesting for business applications that aim to
discover business processes from data. Examples of event logs include
process data generated by administrative services, health care data about
patient handling, and logs of workflow tools. In comparison, quantitative
approaches to model construction are grounded in standard mathematical
representations (e.g., systems of differential equations). Quantitative
representations are common in scientific applications, and are especially
prominent in the environmental and biological sciences that deal with
complex, natural systems. Notably, the business and scientific worlds are
not separated by an interest in the qualitative or quantitative emphasis
of their models. Moreover, researchers working in these domains would
benefit from approaches that integrate the qualitative and quantitative
aspects of system behavior.

In this workshop, we aim to attract researchers with an interest in
inductive process modeling in different formalisms including Petri nets,
qualitative and quantitative processes, differential equations, episode
rules, logical rules, and others. Also, although we have emphasized the
business and scientific domains, we are open to any application of process
model induction. A non-exhaustive list of topics includes:

• learning structured process models such as Petri net or process algebra models from event logs
• modeling techniques for describing the structure of event data such as Markov models
• learning differential equation models
• learning in qualitative reasoning representations
• learning in temporal logic
• learning logical models of state transitions (e.g., by recursive clauses)
• learning from time-stamped event sequences (e.g., episode rules)
• learning from large databases of trajectories
• connectionist/subsymbolic models of sequence learning
• scalable and robust process mining algorithms and techniques
• process mining evaluation: metrics, approaches and frameworks
• the adaption of web mining, text mining, temporal data mining approaches for inductive process modeling
• particularly welcome are case studies and applications (e.g., from business, the environmental, medical or biological sciences) and discussions of the lessons learned from such case studies
• and papers identifying open problems such as dealing with missing and/or noisy data, regularization, incorporating background/domain knowledge, efficient search through the space of candidate process-based models, ...

Inductive process modeling and process mining are challenging research
areas that have the potential to grow in importance like graph or sequence
mining. On the other hand, process mining can benefit from the input of
related fields in data mining and machine learning, such as temporal data
mining, episodes and web log mining. In the ECML/PKDD 2008 workshop on the
induction of process models, we intend to bring scientists together and
actively identify common research threads, define open problems, and
develop collaborative contacts. It should provide a more relaxed
atmosphere than a conference setting where participants are encouraged to
ask clarifying questions throughout the talks and to move past
jargon-induced barriers.

Submission

Extended abstracts (two pages in Springer format) should be submitted by
June 16th, 2008. Final versions of accepted papers will appear in the
informal ECML/PKDD workshop proceedings and will be made available on the
workshop website before the workshop takes place. Submission implies the
willingness of at least one of the authors to register and present the
paper. Authors of accepted abstracts will be asked to submit a short 4 to
8 page paper in PDF format (following the Springer LNCS guidelines for
preparing manuscripts) that describes their research in more detail.

Important Dates

Abstracts due June 16th
Author Notification on June 30th
Final Papers due August 4th
Workshop September 15th

Organizing Committee

Will Bridewell, Stanford University, USA
Toon Calders, Eindhoven University of Technology, The Netherlands
Ana Karla de Medeiros, Eindhoven University of Technology, The Netherlands
Stefan Kramer, Technische Universität München, Germany
Mykola Pechenizkiy, Eindhoven University of Technology, The Netherlands
Ljupco Todorovski, University of Ljubljana, Slovenia