Rules of the selection of descriptors for a model of permuted index

Abstract

A model of a Hungarian-language permuted index was prepared by constructing chains of coordinated free descriptore. The specific problems arising is the source of the preparations may be divided into two groups:
1. problems pertaining to forming descriptors;
2. problems pertaining to forming descriptor-chains.
From the viewpoint of permutation, the basic element has been defined as a single concept consisting of one or more words, being not separated from each other by spacing but by special signs. This facilitated the compression of words, irrelevent by themselves into concepts, as well as the decomposition (by hyphen or other special symbol) of a complex concept the components of which by themselves were, however, relevant.
Each word-chain, composed of free descriptors, represents a logical unit out of the content of the document. It seemed appropriate to organize the model in a form that all word-chains referring to a given document should appear in one group, in subsequent lines. This has been achieved by coordinating to each document, a five-digit number.