Workflow Dependency Analysis and Its Implications on Distributed Workflow Systems

• Authors: Kwang-Hoon Kim
• Venue: AINA '03 Proceedings of the 17th International Conference on Advanced Information Networking and Applications
• Year: 2003

The design and implementation of a workflow management system is typically a large and complex task. Decisions need to be made about the hardware and software platforms, the data structures, the algorithms, and the interconnection of various modules utilized by various users and administrators. These design decisions are further complicated by requirements such as scalability, flexibility, robustness, speed, and usability. In this paper I describe four types of workflow dependency analyses conceived to be helpful in the design of distributed workflow systems, and in the understanding of the spectrum of possibilities for large-scale, domain specific, and flexible distributed workflow architectures. That is, this paper analyzes four types of dependencies, such as data dependency, activity (control) dependency, role dependency, and actor dependency, in a workflow model based upon the information control net. Also, it describes the implications of the dependency analysis mechanisms on architecting an object-based distributed workflow system. In order to perform the dependency analysis, the abstraction (framework) is at first done on a workflow model to generate dependent relationships among data, activities, roles and actors, respectively. Next, based-upon the abstracted workflow model that is formally and graphically represented by four difference types of dependent nets, each of which is generated by its corresponding algorithm conceived in this paper, it is possible to define a series of advanced workflow models and architectures that are appropriately applicable for implementing a distributed workflow system. The main reason why the workflow dependency analysis should be meaningful and valuable work lies just on it.