Character coding for bibliographical record control

Open Access
Publisher Website
Google Scholar
Abstract
The identification and control of bibliographical records is primarily based on a code which usually takes the form of a number arbitrarily allocated to each record. The code accompanies the record throughout its processing history and is normally incorporated into the record as one of its fields. This paper describes a new computer method for generating a Universal Standard Book Code (USBC) comprising alphanumerical characters dependent on the pertinent elements of a bibliographical record such as the Title, Publisher, Date, Language, Edition and Volume. The method is described as Dynamic Associative Coding, where following a statistical analysis based on the frequency count of the characters in certain record fields, the least frequent are selected for inclusion into the code. The universality of the code is ensured by the fact that it does not require any a priori statistical information about the records, it is machine independent, it does not use any look-up tables and the code can be generated at any time without the need to consult a hegemonical body. The method was implemented and tested using over 135,000 MARC (Machine Readable Catalogue) records from three different data bases. The percentage of unique codes generated was found to be well over 99% for each of the files, while in the case of one of the smaller files unique codes were generated for all records.

This publication has 0 references indexed in Scilit: