RELATIONAL DATA FILE: INPUT TECHNIQUES,

Abstract

The document is an input manual for the Relational Data File (see AD-642 120), applicable to any high-volume, structured, data base. A general discussion of problems and requirements precedes the rationale for the procedures selected. Designers of machine-processed files tend to neglect input; nearly always this results in inefficient, costly, and highly error-prone manual preprocessing. Input includes all procedures necessary to file data in machinable form, check, correct, maintain, and update it; this report is concerned with the methodology for the manual extraction and preparation of data for input. The aims are flexibility, generality, and efficiency, using the relative strengths of man and machine in appropriate ways. Synonymity and ambiguity dictionaries are needed by both input workers and file programs. Error correction and detection procedures should be built in. Procedures are given for handling tentative, unknown, projected, or negative data, for ambiguous data, and for relative as well as absolute dating. (Author)

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1969
Accession Number
AD0684121

Entities

People

  • Wade B. Holland

Organizations

  • RAND Corporation

Tags

DTIC Thesaurus Topics

  • Ambiguity
  • Databases
  • Detection
  • Dictionaries
  • Efficiency
  • Extraction
  • Preprocessing
  • Resilience

Readers

  • Computational Linguistics
  • Systems Analysis and Design