The Use of Semantic Structures in Information Systems

An approach to the problem of obtaining a semantically structured vocabulary for use in information systems is presented. It is felt that this approach is valid for a system whose users are a relatively homogeneous group in a well-defined problem area. The semantic structure of the technical vocabulary of such a group is defined and is presented as a useful representation of a relatively stable relationship between the words appearing in queries about information in the system and the words in the information-carrying documents or records themselves. Procedures are given for the discovery of semantic structure and its use in an index to information and for a formal language for describing information.