Job - Based Uncertainty Processing Method For Complex Scene Business Processes

To make business process model adapt to environment change is vital for enterprises and organizations to keep competitive. It is a big challenge to adapt to those changes that cannot be predicated at modeling stage in the complex context.In the complex context, for an activity in a process, the sub-process to deal with each particular situation is known. But when to launch a sub-process and how to compose these sub-processes cannot be specified before the process gets instantiated. Moreover, there exist execution dependencies between sub-processes. Existing methods cannot deal with these problems well in such a context.In this article, we present a method based on ontology to deal with the uncertainty of business process at runtime in the complex context. In the method, the changes that can be predicated at modeling stage are specified using loosely specified processes; those changes that can’t be predicated at modeling stage are represented by empty activities that have no concrete operations specified. The specifying of concrete operations is deferred to runtime. This deferred specifying mechanism is composed of two parts:the situation-aware process composition method and dependence handling between sub-processes.In the situation-aware process composition method, the sub-processes are driven by situations in the domain. In this method, ontology is used to describe the entities, properties and relations between them. A property of an entity could take different values. A combination of properties with different values in an activity corresponds to the hold of a situation. The sub-processes linked to those situations that hold are executed after user selection.There are dependencies between sub-processes due to business, policy, data, etc. That means before a sub-process is executed, another sub-process, the depended sub-process, must be executed and completed. According to all or only one of the depended sub-processes need to be executed, AND type and OR type dependence relation is distinguished. We use the model based on DAG to model the dependency relation. According the model’s property, an algorithm to determine the dependency satisfaction of an activity is presented. The dependency is also captured using ontology. Use the method in this article, the models created is easy to be shared, be used, and be understood. The implicit knowledge in the model can be inferred and the consistency of the model can be checked.