Databases Reference
In-Depth Information
vices by accommodating extended order tracking capabilities, including both the internal
order handling process and external product delivery process. To do so, the online store is
to fi rst integrate the two local process models into one inter-organizational process model.
If the local process models and their interactions are specifi ed explicitly and coordinated
by one organization, it is possible to integrate the two local process models (van der Aalst,
1999). However, such an assumption of full information sharing is hardly accepted in the
modern business environment because of the autonomy problem stated in section 1. For
example, when the transportation company does not want its operational mistakes to be
known to customers, it wants to hide the process information related to A
2,6
(reimburse-
ment for the damaged product). Thus, the transportation company should construct a partial
process model by removing A
2,6
from its product delivery process PM
2,
so that the online
store uses the partial process model instead of full PM
2
to compose an inter-organizational
process model. However, such a process information sharing scheme that supports sharing
policies among participants has been rarely dealt with in the existing research on inter-or-
ganizational processes.
The process information sharing system of the online store constantly captures and
merges the process statuses of the order handling and product delivery processes in order to
provide customers the dynamically changing global process status. In doing so, the system
refers to the inter-organizational process model, including the two process models, PM
1
and
PM
2
, and their interactions at the run time, and thus the details of the inter-organizational
process model should be encoded in the system. However, achieving agility in companies
often causes a change of a process model or sharing policy. When such a change arises, the
system should be re-implemented and newly created to refl ect the change with maintenance
cost and time.
The following sections propose the federated process framework and show how the
framework overcomes these problems while facilitating effective implementation of process
information sharing.
FEDERATED PROCESS FRAMEWORK
The federated database system approach is referred to in the development of the feder-
ated process framework. A federated database system was originally proposed to facilitate
information sharing among cooperating but autonomous local database systems (Heimbigner,
1985; Sheth, 1990). While the local database systems independently perform local opera-
tions, they only provide partial and controlled data for the requests of the federated database
system. Typical development process of the federated database system is composed of the
following four steps:
(i) Standardizing local database schemas;
(ii) Controlling and restricting the schemas;
(iii) Integrating the multiple schemas into one schema;
(iv) Customizing the integrated schema for end users' needs.
In this chapter, the federated database approach is adopted and extended as a framework
for the process information sharing among collaborating but autonomous business processes.
Specifi cally, this section proposes the four steps of the federated process framework in detail.
