A language for implementing information retrieval software

Information storage and retrieval (IS&R) programming is often hindered by the lack of specialized facilities in traditional computer languages. This article describes a portable programming language, derived from the C and Mumps languages and designed to meet the needs of IS&R applications. The language allows the user to deal with an IS&R database at a level of abstraction consistent with many theoretical models commonly in use. Examples from a pilot implementation of a statistically‐based retrieval system are given which demonstrate that this approach results in substantial savings in disk storage and greater coding efficiency when compared with traditional implementation techniques.