E-services usually span multiple organizations and evolve in time-domain due to the changes in the environment pertaining to each organization. Enacting composite e-services over evolving inter-organizational business processes is a challenging task. In this paper, we highlight modeling of business processes for eservices and present enactment of the e-services with the help of meta models in order to cope with the changing needs. We also describe high-level framework to enact e-services through actionable meta models.
Modeling e-services is an important task in business processes specifications and execution. A single model may not be enough to design e-services when it spans multiple organizations. At the same time, it is hard to describe a generic model for any kind of eservices. This necessitates an abstract level of modeling, that is meta-modeling, which helps in instantiating context-specific model instances to suit the needs of e-services. It also provides required facilities to support functionality for adapting the model to new requirements.
In the domain of business process modeling, meta-models take a leadership role in
defining how the data and corresponding application/business logic is handled right from
the modeling to all the way to enactment of business processes. E-services involve
interorganizational business processes. Enactment of such business processes involves multiple
parties and is modeled, using workflows by introducing e-services in business process
models [
Task a1 Task a2 Task a3 Task a4 Task b1 Task c1 Task c3 Task b2 Task c2
framework. Rosemann and Muehlen [7] presented a meta model approach to evaluate and compare different workflow management systems. They used extended entity relationship models to design the meta data models and provided a generic organizational reference meta model for enacting inter-organizational workflows. Design of all these meta models are specific to an application and assumes complete specifications of it is known in advance.
In this work, we consider services as workflows and present how meta-models are useful to facilitate the changes in specifications during run-time.
Actionable meta modeling enables upkeep of meta model and the corresponding mapping all the way to run-time caused by the changes in organizational environment at run time, or sometimes changes initiated by users due to external factors. The main feature of this framework is that it facilitates to systematically address the problem of proactively responding to changes driven from environment.
(Run-time Environment)
Reconciliation
During e-service system design, the requirements are collected and the e-service elements such as activities and payments are extracted to model an e-service. Workflows will be identified to execute the e-service activities carried out by organizations. The log records keep track of the e-service execution. However, changes in organizational environment during e-service enactment need modification at conceptual level as well as at logical level. In order to propagate changes that occur during enactment to conceptual and logical level, there is a need to obtain business logic from log records and should be reconciled with the actual business logic of the e-service (Fig. 2).
In the next section, we present the framework for enactment of e-services for evolving inter-organizational business processes.
Architecture for Actionable Meta-modeling E-service execution requires the modification of model schema definitions and in-turn changes in workflow instances, as a remedy. Moreover, additional constructs are needed in the system during work in progress. The present approach offers a practical solution from modeling and enactment of e-services, driven by meta-models. We maintain business policies that refer to adaptable instances of workflow schema while the service is being executed. Workflow patterns [4, 8] are useful to describe how the changes will be specified, implemented and perceived in e-services.
Figure 3 shows architecture of a system to model business processes and its instances. Web Service Server provides the services and transport of inter-organizational communications among business partners involved in the service. The Run-time environment (RTE) details such as workflows, rules, etc. are maintained in the database. Dynamic Workflow Instance Generation (DWIG) generates workflows on-the-fly and rules. It also allows the administrators to customize and edit them. Workflow definitions created or specified are executed by the Meta-workflow Driven Workflow execution engine. That is, the workflow engine enacts the workflows specified by the dynamic workflow instance Generation. The Event handler manages the events occurring during the execution of workflows. It handles events in a unified manner for both normal and exception parts of a business process workflow. The Event-Condition-Action (ECA) Rule Manager initiates appropriate ECA rules based on the input from Event Handler. It also keeps track of generated rules with their corresponding actions and allows users to define additional rules if necessary.
The workflow engine and the ECA rule manager works in a synchronized manner. Thus, the ECA rules control the workflow execution and the events that occur during the workflow execution result in appropriate actions. The changes in the design-time update the corresponding database. The Knowledge Base maintains the currency of information such as application policies, versions, new requirements etc. that governs execution of the workflow. It also stores the vocabulary of an e-service domain. Knowledge Base captures the updates that take place in the run-time as well as in the design-time. These updates, in turn, become input to the Run-time Business Processes evaluator (RTBPE), which generates candidate meta-model(s) and add/modify the meta-model repository. MetaModel repository contains the meta-models that are specific to a service under consideration. Meta-models are added or modified based on the requirements collected from run-time environment changes as well as design-time changes.
Business Processes logic and events are captured by execution support of metamodels. Control and Monitoring Unit (CMU) generates/modifies the workflow patterns and also specific elements namely Events, States, Processes and Rules (ESPR) in order to keep track of changes and their progression with respect to business processes under consideration. CMU receives events and results from the run-time environment and service specific components and policies. The modeling of changes during service Public UDDI Register
Run Time Business Processes Evaluator
Control and
Monitoring Unit
Meta-Model Selector Knowledge Base
Run Time Environment
Database for workflows, Rules, etc.
Event Handler ECA Rule Manager enactment can be seen as a different kind of meta-processes (tasks). Further, Service Specific Components are required for encapsulation and realization of the domain-specific logic for the service. The three components namely CMU, workflow patterns and ESPR will serve as RTBPE. Meta-model selector selects the appropriate meta-model(s) from the meta-model repository, which in turn drives the dynamic workflows to carry out business processes execution according to the changed context.
Discussion on the Proposed Meta-modeling Architecture Proposed architecture facilitates learning from Business Activity Monitoring (BAM) to feed into the meta model updates. In the current Business Process Management domain, there is a lack of suitable architectures for BAM to support learning. The RTBPE proposed in this architecture provides the BAM metrics and their respective evaluations. These evaluations are the primary drivers for the learning required to update the meta models. The meta model learning is fed back into RTE through the dynamic work flow link with meta model repository. The benefits derivable from the proposed architecture depend on the e-service BAM design in terms of the RTBPE. Another advantage is that the dynamic workflow link between the meta model repository and RTE makes this architecture actionable.
Further work is needed to enhance the proposed architecture by explicitly specifying the service design issues such as granularity, complexity, flexibility, interoperability, autonomy, contract specifications, etc. in the RTBPE. Such enhancements will help the Business Process Management Systems (BPMS) software developers to bench mark their products and the BPMS users to assess and evaluate the BPMS products in a more comprehensive manner. 4
Modeling of e-services facilitates effective specification and execution of business processes. In the case of inter-organizational business processes, development of a generic model is a complex task. In this paper, we presented a meta modeling approach which helps to create context based model instances according to the changes that occur during run-time. Moreover, the meta-models learn from the services environment and update the models as necessary and hence making the meta-models actionable.