Semantic structures of statistical meta-data

Abstract

A re-definition of the concept of statistical meta-data can be based on the concept of ‘statistical indicator’. The structural (semiotic) model of ‘statistical indicator’ is seen as the fundamental semantic structure of statistical meta-data. Relations between statistical indicators (i.e. partitive, procedural, hierarchical, lexical, associative and attributive) also merit discussion. The main structural forms of meta-data can be developed and specified on this basis: classification, nomenclature, systematized list and descriptor language. These structural forms of meta-data are used for the design of statistical surveys and the development and maintenance of registers and dictionaries, which are the instruments for the realization of meta-data functions in the statistical information system: interpretation, identification, organization and co-ordination of data and surveys.