Databases Reference
In-Depth Information
As advances in information systems take place to facilitate business trans-
actions, organizations are seeking ways to effectively integrate and automate
their business processes. The advent of database technology has made the
change of data more adaptive by successfully separating the access of data
from the applications. However, any change and enhancement to the business
policies would entail modifying application codes, as the business policy is still
often hard-coded in applications rather than accessible to all systems. Work-
flow systems are a step in the direction of providing both automation and
reengineering functionalities. The fundamental idea of workflow technology is
to separate the business policy from the business applications to enhance flex-
ibility and maintainability of business process reengineering. This separation
facilitates reengineering at the organizational level without delving into the ap-
plication details. Other advantages include supporting resource allocation and
dynamically adapting to workload changes. As a testament to the recognition
of these benefits, workflow systems are today used in numerous business ap-
plication domains including oce automation, finance and banking, software
development, healthcare, telecommunications, manufacturing and production,
and scientific research.
Workflow management aims at modeling and controlling the execution of
business processes involving a combination of manual and automated activi-
ties in an organization. A workflow is defined as a set of coordinated activities
that achieves a common business objective [1]. Thus, a workflow separates the
various activities of a given organizational process into a set of well-defined
activities, called
tasks
. A task is a described piece of work that contributes
toward the accomplishment of a process [23, 15]. Tasks may be carried out
by humans, application programs, or processing entities according to the or-
ganizational rules relevant to the process represented by the workflow. Tasks
that build up the workflow are usually related and dependent upon one an-
other, which in turn are specified by a set of execution constraints called task
dependencies. These task dependencies play a key role in supporting various
workflow specifications such as concurrency, serialization, exclusion, alterna-
tion, compensation and so on. To ensure the correctness of workflow execution,
tasks need to be executed in a coordinated manner based on these dependency
requirements. A workflow management system (WFMS) is a system that sup-
ports process specification, enactment, monitoring, coordination, and admin-
istration of workflow process through the execution of software, whose order
of execution is based on the workflow logic [1]. In the following, we provide
an example workflow to facilitate understanding of tasks and dependencies.
Example 1.
Consider a travel reimbursement processing workflow [2] as shown
in Figure 1. This workflow consists of four tasks: preparing a claim (
T
1
),
approving the claim (
T
2
), issuing a check (
T
3
) and notifying the employee in
case the claim is denied (
T
4
).
Coordinating constraints between the tasks are represented by dependen-
cies shown above the arrows connecting the tasks. The task dependency “bs”
