MDA-Based Interoperability Establishment Using Language Independent Information Models [chapter]

Carlos Agostinho, Jaroslav Černý, Ricardo Jardim-Goncalves
2012 Lecture Notes in Business Information Processing  
Nowadays, more and more enterprises realize that one important step to success in their business is to create new and innovative products. Many times the solution to do that is to abandon the idea of an enterprise as an "isolated island", and get collaboration with others: worldwide non-hierarchical networks are characterized by collaboration and non-centralized decision making. This paper proposes a conceptual model common to the entire business network, in a framework that enables the
more » ... ion of individual models at their meta-level and increase language independency and interoperability, keeping all the enterprise software's integrity intact. The strategy presented allows an incremental mapping construction, to achieve growing integration. Introduction Interoperability is a property directly related with the heterogeneity of model languages, communication capabilities, databases and semantics. Differences in these hide a great barrier to achieve the time-to-market symbiosis that can unleash a solution more valuable than the sum of its creators. Interoperability is more than just a communication support: it is a software approach to maximize the benefits of diversity, rather than to integrate the different system into one. Such diversity leads to more fruitful results than by just integrating different systems into one. Since many organizations developed and purchased software solutions based on their own needs, the required cooperation with others is not a trivial activity and business partnerships are less effective, evidencing low level of interoperability. To solve this problem, instead of adopting a paradigm that obligates every organization to migrate their systems, or develop complex mappings in a single step to comply with these advanced practices, one can act at the communication module, where the data is exchanged. The authors propose Model Driven Architecture (MDA) based technologies for the development of transformations and execution of automatic and executable Model Morphisms (MoMo), also providing traceability and repeatability on them. The proposed framework enables to respond automatically to the network dynamics and its sustainability, i.e. changes that occur over the time and impact negatively the interoperable state can be tuned and balanced. Model Driven Engineering Model-Driven Engineering (MDE), sometimes also referred as Model-Driven Development (MDD), is an emerging practice for developing model driven applications. It represents a promising software engineering approach to address systems complexity, both by simplifying and formalizing the various activities and tasks that comprise an information system life cycle (i.e. from design, to construction, deployment, operation, maintenance and modification). Given today's increase of technology complexity, models are becoming a powerful mechanism to precisely describe problems in a way that avoids delving into technological details, thus allowing developers to focus on more abstract tasks and increasing productivity rather than to computing concepts. MDE is meant to maximize compatibility between systems, simplifying the process of design, and promoting communication between individuals and teams working on the system [1]. MDD/MDE's vision goes even further, invoking the unification principle, which states that "everything is a model" (i.e., platforms, components, legacy software, services, etc.), encouraging the support of models at different levels of abstraction, from high-level business models focusing on goals, roles and responsibilities down to detailed use-case and scenario models for business execution [2, 3] . These models are developed through extensive communication among product managers, designers, and members of the development team, and as they approach completion, enable a fast development of product and systems. However, despite obvious potential capabilities for closely matching the EI holistic levels, yielding major productivity and reliability benefits, there is not yet consensus about its technology readiness [4, 5] . Level 0 (Data) Level 1 (Information Models) Level 2 (Meta-models) Level 3 (Meta-metamodels)
doi:10.1007/978-3-642-33068-1_14 fatcat:t57uiwm6cjesbdw34hfovknys4