Abstract
Purpose - The purpose of this paper is to show that exploiting fundamental ideas of granular computing can lead to further conceptual developments of granular metastructures, which are inherently associated with computing involving a large number of individual datasets; and to show that such processing leads to the representatives of information granules and granular models in the form of metastructures and metamodels.Design methodology approach - The formulation of the concept of granular metastructures is provided and presented along with some essential algorithmic developments and associated optimization strategies. The overall methodological framework is the one of granular computing, especially fuzzy sets and fuzzy sets of higher type. Given the structural facet of optimization, the paper stresses the relevance of the use of evolutionary optimization.Findings - This paper focused on the underlying concepts and while it elaborated on some development aspects and optimization tools, it should be stressed that further refinement and a thorough exploitation of optimization techniques in application to the inherently combinatorial facet of the problem are to be pursued in detail.Practical implications - The introduced approach and algorithms could be of interest when solving problems of granular metastructures, in particular those encountered in knowledge-based systems.Originality value - The main aspects of originality concern a formulation of the concept of granular metastructures and their design, based on granular evidence (experimental data) of lower type. A constructive way of forming type-2 fuzzy sets via the principle of justifiable granularity exhibits a significant level of originality and offers a general way of designing information granules.