Ontology driven software engineering: models, methods, implementations

Ontology-driven software engineering is the scope of the paper. A retrospective analysis of the approaches, models and methods is carried out. In the first section, a brief overview of "classical" technologies, approaches and methods for software engineering is presented. It is shown that the emphasis of research and development is currently shifting from the field of technologies of programming towards the modeling of the software design processes, as well as towards the specifications and testing of the developed systems. It is outlined that the general trend in the domain is the use of methods and tools of ontology modeling of both the design processes per se and the specifications of the developed systems. The second section is devoted to a review of publications in the field of ontology-based software engineering. Such paradigms for modeling software development processes as Software Product Lines, plug-and-play software development - Component-Based Development, model-driven software engineering - Model-Driven Engineering, etc. are discussed. It is shown that “programming through generation”, i.e. Generative Programming, is a general approach supporting the considered paradigms. It is shown that R&D in the domain of the semantic models of software engineering is currently focused on a new approach - Ontology-Based (-Driven) Software Engineering. Within the framework of this area, ontology engineering tools are developed along with modeling the software development processes. The scientific and technical perspectives of ontology driven software engineering are specified and discussed in the final part of the paper. In particular, it is outlined that the theoretical foundations of ontology-based modeling of software design processes form the solid basis for shifting the focus from the models of compiling the modules to their pattern-based generation from internal specifications. It is noted that the next trend in the field of OBSE-ODSE will be the usage of machine learning methods and tools to generate ontological models of software design.