Some Results of Statistical Studies of a Descriptor Language

Abstract

The statistical regularities of a descriptor language are investigated. The distribution patterns of descriptor occurrence in document and request search patterns are examined. The distributions are shown to be close to the Zipf-Mandelbrot law. The statistical proximity of the frequency lists of descriptors from document and request files was measured. The dependence was evaluated of descriptor frequency on such characteristics as the number of keywords in the equivalence class of the given descriptor, the number of broader term references to the given descriptor from its underlying ones, and the average frequency of occurrence of the keywords. The validity of the findings was estimated.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 30, 1971
Accession Number
AD0734906

Entities

People

  • V. K. Vakhabov

Organizations

  • National Air and Space Intelligence Center

Tags

DTIC Thesaurus Topics

  • Accuracy
  • Agreements
  • Air Force
  • Coefficients
  • Computer Languages
  • Dictionaries
  • Errors
  • Foreign Technology
  • Frequency
  • Information Retrieval
  • Language
  • Linguistics
  • Machine Translation
  • Natural Languages
  • Probability
  • Reliability
  • Test And Evaluation

Readers

  • Computational Linguistics
  • Statistical inference.