Dynamic Model Harmonization Between Unknown eBusiness Systems

Abstract

It is usually assumed that systems share an agreed model of business protocol to implement consistent exchange of series of business messages between eBusiness processes. R&Ds based on this approach are making progress such as model repository technology, business process modeling and business process description languages. This assumption, however, is not always satisfied when considering flexible and free dealings among independent and autonomous eBusiness systems demanded in future eBusiness environment, where systems having their own way of feasible interactions without an agreed business protocol model unexpectedly encounter each other, and want to execute business conversation using their best efforts. We propose a solution to the above challenge in this paper, taking an approach: Each system exposes a model specifying possible interfaces and behavior; Systems exchange exposed models; Each system reduces its exposed model adjusting to the opponent system; Then, systems begin business conversation if both models were successfully reduced. The paper provides a formal definition of exposed models and an algorithm to reduce exposed models, and presents results using experimental implementation showing appropriateness of the proposed definition and algorithm. It also discusses matching methods of business resource definitions used in the algorithm. The results firstly target eBusiness application but are also applicable to dynamic protocol generations in other areas including robots, mobile devices and ubiquitous environments.