Business Modeling and Component Mining Based on Rough Set Theory

Model based approach and component based software development (CBSD) both contribute to software reuse. In this paper, we present a formal approach to combine them together for efficient software reuse. The approach consists of two phases. The first is the model construction phase. Rough set theory (RST) and colored Petri nets (CPN) are used in order to build the accurate model (hereafter, simply referred to “the model”) from various facts and requirements provided by various domain-experts. RST coordinates differences in knowledge and concepts among those experts, whereas CPN express the model rigorously and intuitively. The second is the component mining phase. We use Σ algebra and RST for retrieving such components as being adaptable to the model. Σ algebra evaluates functional equivalency between the model and the components, while RST coordinates differences in sorts or data types between them. We mainly focus on large scale enterprise back-office applications, however the approach can be easily extended to the other domains, since it does not depend on any specific domain knowledge. An application model example from an order processing is depicted to show how our approach works effectively.