Report on HICSS-35
Minitrack
Integrated Modeling
of Distributed Software Systems
and Workflow Applications
Mathias Weske, HPI Potsdam,
Germany, weske@hpi.uni-potsdam.de
Guido Wirtz, University of
Münster, Germany, guidow@math.uni-muenster.de
In this year’s Minitrack on
Integrated Modeling of Distributed
Software Systems and Workflow
Applications at the Software
Technology Track at the 35th
Hawaii International Conference
on System Sciences, researchers
and practitioners from Asia,
Europe, North America and South
America gathered to discuss
research topics of practical
relevance in the field of process
technology and distributed systems
technology. This year’s program
consists of six contributions,
organized in two sessions. The
contributions cover a variety
of issues ranging from the analysis
of workflow process specifications
to case studies in applying
distributed workflow techniques
to specific application areas.
In
the first session, issues related
to the specification of workflows
and their analysis are discussed.
Lin, Zhao, Li, and Chen from
the Peoples Republic of China
investigate synchronization
properties and deadlocks in
workflow graphs (H. Lin, Z.
Zhao, H. Li, and Z. Chen: A
Novel Graph Reduction Algorithm
to Identify Structural Conflicts).
This contribution introduces
a new graph reduction technique
that reduces structural correct
graphs and, hence, reveals errors
in the workflow process specification
if the reduction does not succeed.
Inoue and Iwaihara from Japan
discuss a method to handle exceptional
situations and dynamics of systems
during the design and implementation
process (S. Inoue and M. Iwaihara:
Supporting Dynamic Process Specifications
Using Communication-Based Processes).
The method assumes a communicative
process as the driving force
during the realization of a
system and supports processes
that are not completely known
at design time. The contribution
reports on a prototypical system,
called M-Trans, that is based
on a notion of consistency between
so-called templates and their
implementation. Kwak, Han, and
Shim from South Korea put their
focus on the specific problems
specific to interorganizational
workflow systems (M. Kwak, D.
Han, and J. Shim: A Framework
Supporting Dynamic Workflow
Interoperation and Enterprise
Application Integration). A
framework to support interoperation
between workflow management
systems is proposed, and the
applicability of that approach
to Enterprise Application Integration
is sketched. Especially, the
heterogeneity of the involved
workflow management systems
raises some important questions
when discussed in a dynamic
context, i.e., without specifying
all details of possible interconnected
systems a priori. A framework
that is able to handle these
problems is presented and evaluated
in the context of the parts
inventory management process
of a motor company. The work
is based on workflow standards
as proposed by the Workflow
Management Coalition (WfMC)
as well as current XML-based
approaches for workflow systems.
In the second session, comparative analysis of
recent work and the integration
of methods from different areas
of research in workflow management
and distributed systems are
a common characteristics of
the papers presented. Breton
and Bézivin from France describe
two recent standards proposals
by the WfMC and the OMG for
workflow management and compare
them on the basis of a proposed
meta-model regarding workflow
dynamics (E. Breton and J. Bézivin:
Weaving Definition and Execution
Aspects of Process Meta-models).
The paper provides an in-depth
state-of-the-art report with
a focus on dynamic aspects,
which are highly relevant in
the workflow context. A contribution
by Bastos, Dubugras, and Ruiz
from Brazil discuss an extension
of activity diagrams taken from
the Unified Modeling Language
to model workflows (R. Bastos,
D. Dubugras, and A. Ruiz: Extending
UML Activity Diagram for Workflow
Modeling in Production Systems).
In particular, based on the
C-WF model that combines process
modeling capabilities of the
CIMOSA and WfMC approaches,
Workflow Activity Diagrams are
introduced. The close relationship
between issues in engineering
complex distributed software
systems and in engineering workflow
systems is focused in this contribution.
Leon Zhao from the United States
uses a case study from the area
of document flow to investigate
the relationships between issues
from knowledge management, knowledge
distribution and coordinated
document flow (J.L. Zhao: Workflow-Centric
Distribution of Org-anizational
Knowledge: The Case of Document
Flow Coordination). The combined
use of these technologies leads
to a promising concept of organizational
knowledge networks, which is
presented with its theoretical
background, i.e., knowledge
association algebra. The work
presented shows that techniques
and lessons learned from the
different areas are of mutual
benefit, but serious problems
have been identified, too. Among
them are the consistent incorporation
of all aspects relevant to workflow
and practically usable methods
to combine the dynamic aspects
of workflows with the well-known
static structure when modeling
for a distributed context. Finally we mention that the full text of the papers discussed
here as well as all other HICSS
conference papers since 1999
can be downloaded from www.hicss.org by selecting the HICSS Digital
Library Link, a service that
we very much appreciate. |
