Methodology for rapid development of PAIS based on ontological modeling and category theory
TL;DR
This paper proposes a methodology for the rapid development of Process-Aware Information Systems (PAIS) based on ontological modeling and category theory, which distinguishes the categorical specification, in which functoriality follows by construction, from its operational implementation, whose conformance is established through explicit engineering checks.
Abstract
Model-Driven Architecture (MDA) structures software development through computation-independent, platform-independent, and platform-specific models. Its practical value depends on machine-readable models and transformations that preserve relevant structural properties while producing executable code. This paper proposes a methodology for the rapid development of Process-Aware Information Systems (PAIS) based on ontological modeling and category theory. Ontology- Driven Code Generation (ODCG) is formalized as a deterministic, multi-phase transformation from the BCore ontology, through a platform-specific ontology and library, to source code. Model levels are represented as categories, while transformations are represented as functors. The main contribution is a categorical formalization of stack-parametrized multi-target generation, together with a conditional guarantee of schema-level structural preservation under explicitly stated verification assumptions, including mapping completeness, injectivity, quotient compatibility, and absence of unintended collapse. The methodology distinguishes the categorical specification, in which functoriality follows by construction, from its operational implementation, whose conformance is established through explicit engineering checks. The approach is implemented and evaluated for the Java platform through the BABj ontology and library by generating code for an operational business information system during its life cycle.