Dual Deep Modeling of Business Processes

Abstract

Multi-level modeling (MLM) facilitates conceptual modeling at multiple levels, with clabjects as basic modeling constructs that combine characteristics of metaclasses, classes and objects. Different MLM approaches differ, among others, in the meaning and structure of levels and clabjects, in the strictness or flexibility regarding cross-level relationships, and in the mechanisms for deep characterization by which clabjects at higher levels describe and constrain clabjects at multiple lower levels. The Multi-level Process Challenge provides a testbed for MLM approaches to highlight design decisions regarding these aspects. In this paper we solve the challenge using Dual Deep Modeling (DDM), a MLM approach that features dual potencies which facilitate high flexibility for cross-level relationships. With relationships with dual potencies, a single clabject can play multiple roles at different levels of instantiation, thereby DDM facilitates very compact multi-level models.