For large meta information of national integrated statistics

  • Authors:
  • Hideto Sato;Ryosuke Hotaka

  • Affiliations:
  • -;-

  • Venue:
  • SSDBM'81 Proceedings of the 1st LBL Workshop on Statistical database management
  • Year:
  • 1981

Quantified Score

Hi-index 0.00

Visualization

Abstract

Integrated statistics, synthesized from many survey statistics, form an important part of government statistics. Its typical example is the System of National Accounts. To develop such a system, it is necessary to make consistent preparation of 1) documents of methods, 2) programs, and 3) a database. However, it is usually not easy because of the large amount of data types connected with the system. In this paper, we formulate a language as a means to supporting the design of statistical data integration. This language is based on the data abstraction model and treats four types of semantic hierarchies; generalization, derivation, association (aggregation) and classification. We demonstrate that this language leads to natural documentation of statistical data integration, and meta information, used in both programs and a database for the integration, can be generated from the documents.