INSTRUCT: a teaching package for experimental methods in information retrieval. Part II. Computational aspects

Abstract
INSTRUCT is an interactive computer program illustrating some of the techniques which have been suggested for use in future text retrieval systems. The program is designed as a computerised teaching aid for students in schools of librarianship and information science. A previous paper has provided a general overview of the novel retrieval facilities provided by INSTRUCT. The present paper discusses the means by which these facilities have been implemented: in particular, detailed descriptions are given of the nearest‐neighbour search algorithm that underlies the string similarity and ranked output modules of the program, and of the term‐weighting schemes that are used. The paper concludes with a discussion of potential enhancements to the present program.