Library and Information Science

This paper describes an evaluation of minicomputer-based programming languages from the standpoint of information retrieval technology. Three clinical record retrieval system, namely, a kidney dialysis patient medical system using TOD (time-oriented database) method, a celebral stroke information retrieval system, and a diagnosis-oriented summary system using POS (problem oriented medical system), are shown as the evaluation sample.

At the design stage of these application systems, data elements, data attributes, and data structures are defined and several common functions and algorithms such as string manipulation, tree-structured database filing, logical operation and online feature are extracted.

Secondly, at the implementation stage, the selection and evaluation of the programming languages required for each clinical retrieval application are discussed. The check list chosen from the functions and algorithms is proposed. And such high level languages as COBOL, FORTRAN, APL, MUMPS, and BASIC are compared using the check list. MUMPS is selected as the best suitable language for clinical record retrieval programming.

This type of consideration plays a major role in the design and development of information retrieval systems.