Methodology of reversible generalization in context of classification of information transformations

An approach to the problem of creating a methodology for the reversible generalization of information objects (RGIO) as the processes of their concrete-abstract transformations is proposed. From the theoretical point of view, the description of the processes of generalization of information objects makes it possible to enrich the classification of information transformations in the domain of informatics by including a class of concrete-abstract transformations into it. It is important to emphasize that in the paper, informatics is treated according to the scientific paradigm of Paul Rosenbloom as a complex of scientific disciplines that studies information transformations in technical, living, and social systems, and not just in computers and networks. From the applied point of view, the description of generalization processes in the Rosenbloom paradigm enriches the foundations of developing information systems that implement concrete-abstract transformations. The description of the problem of creating the RGIO-methodology is illustrated by the processes of generalization of bilingual annotations in a supracorpora database (SCDB). The reversibility of the processes of annotation generalization is the basis for a multifaceted and verifiable statistical analysis of arrays of annotations formed in the SCDB.