Programme system for thesaurus compilation written with PL/1 programme language

Abstract

Description of a programme system for mechanizing eseential parts of formally manual operations in thesaurus compilation. The first chapter includes the definitions of the structural units and concept relations most common in thesauri. These are:
hierarchical relation: generic and specific
preferential relation: "see" reference, forbidden,reference, association trails.
The programme system handles all these, further any other relations occuring during application.Thesaurus compilation as well as maintenance of thesaurue files stored on magnetic discs could be made possible by the programme system in two directions:
elaboration and maintenance of subject heading lists complemented with concept codes or source indications: on this base KWIC or other permuted index types as well as subject heading indexes by codes are producible according to request;
terms belonging to subject heading lists could be inoluded into a thesaurus arranged in pairs, linking with relation operators and attached with concept codes: programmes work out the structured part of the thesaurus, produce the term-file as well as the alphabetical EHIC or the concept-code indexes.
Another advantaga of the programme-system that terms could be segmented in compound words, parts with individual content meanings could be denoted by terminators and so they are represented in many places in all the indexes /according to all components/. Method for the estimation of programme time is described, detailed quantitative and time data of experimental processing are given.