In information systems based on a multidimensional approach, a data model is a multidimensional data cube. If one uses a large set of aspects for the analysis of data domain the data cubes are characterized by substantial sparseness. This makes it difficult to describe the metadata of the information system and complicates the organization of data storage. To describe the structure of a sparse data cube, a cluster method can be used. This method is based on the construction of groups of members which are semantically connected with other groups of members. Connected groups related to different dimensions describe the cluster of cells. Classification schemes that correspond to the structural components of the observed phenomenon can be used to describe it’s semantics. Every classification scheme is a graph describing the hierarchy of members that are associated with a separate structural component of the observed phenomenon. The coupling between several classification schemes related to different structural components helps to describe the metadata of the multidimensional information system. Classification schemes are a source of classification of information objects of a multidimensional cube related to the structural components of the observed phenomenon. Copyright © 2018 for the individual papers by the papers’ authors.