A Study on Compiling Thesaurus in Use for Information Retrieval

情報檢索에 必要한 Thesaurus 작성요령

This study deals with compiling thesaurus in use for information retrieval. In compiling thesaurus by means of manual system, there are given four categories of collecting vocabularies, selecting keywords, defining correlation of keywords and compiling them. in compiling thesarus by computers, (1) in order to select keywords from words collection automatically, occurance-frepuency, distribution and importance factors of the words gave been fully considered; (2) in order to process relations of keywords, co-occurence has been taken into consideration; (3) in order to define hierachical aspects of keyword, associative matrix [C] between literatire and keywords had been obtained, and finally the extent of factors of [S] has been taken into consideration: (4) with regard to synonyms of keywords, relative coefficient between keywords has been derved according to the list of the relative coefficient, a relative profile has been fixed, and finally thes profile has been applied to the linear associative information retrieval system.

오늘날 情報의 蓄積 檢索方法中에서 가장 널리 이용되는 것은 Keyword 에 의한 것으로, 이 方法에는 檢索의 효율을 높이기 위하여 各 Keyword의 同義 關係, 階層關係, 同形異義關係 등을 적절히 調節하여 槪念을 聯合할 수 있는 Keyword list가 必要하며, 이 list 를 thesaurus라 부른다. thesaurus는 電子計算 機에 의한 自動情報處理에 없어서는 안될 重要 道具(tool)로서 世界各國에서는 여 러 분야의 것을 작성, 使用하여오고 있다.