Materials Data Typology

Abstract

Technologies for storing and processing vast amounts of data have opened a new stage in the development of materials science, based on the application of artificial intelligence methods to the results of many years of research. Large volumes of heterogeneous data combined with powerful analytic facilities have allowed us to significantly expand the range and rate of production of research in comparison with empirical methods of selecting materials with specified properties. The emphasis is placed on the specifics of these data, which mainly determines the level and capabilities of information technologies in modern materials science. Their main features are revealed, which guarantee sufficient completeness of the information needed for creating and using various materials. These features include coverage, along with properties, of data on the microstructure and technology of the material; a large amount of qualitative information; and semistructured data type, i.e., the absence of a regular presentation scheme. Using the example of a number of infrastructure projects, the potential of management of materials science data taking into account their volume, logical structure and format are considered.