The paper provides a brief overview of the HyperMeData language specifically designed to support data interchange among heterogeneous information systems, and pays attention to meta-level transformation descriptions. The language is sufficiently complex and powerful to catch both intra- and inter- data schema relationships (i.e., to describe both data schemas and data transformations), nonetheless, its routine use requires employing a set of meta-level transformation rules to handle typical schematic differences among semantically similar database objects. The classification of schematic heterogeneities in multidatabases is used as the basis for proposing inter-attribute correspondences (meta-level transformation rules) and the respective translations to HyperMeData descriptions of data transformations (transformation rules).