Design of Controlled Vocabularies

Language is the chief embodiment of information and it is, therefore, not surprising that in the information age, a new importance should attach to language. The 1980s have been called ‘‘the decade of language.’’ Worldwide the present publication of language tools, such as dictionaries, glossaries, and thesauri, is proceeding at the rate of 500 per month and there now exist over 10,000 vocabulary standards in 35 languages. Controlled vocabularies (CV) form an important class of language tools. A vocabulary is a ‘‘list or collection of terms or codes available for use (as in an indexing system).’’ A vocabulary is said to be controlled if it consists of a restricted subset of possible terms. Such a subset, in that it contains only those terms‘‘authorized’’ for use, is sometimes called an authority list. In addition to terminological restriction, most CVs articulate semantic relationships between terms in the vocabulary, the most common of these being the inclusion of hierarchical relationship. Controlled vocabularies can take many forms and be used for many purposes. In the form of terminological databanks, CVs are used to assist in both manual and automatic translation. In the form of glossaries, they standardize and explicate the meaning or usage of terms in specialized fields of activity. In the form of literary thesauri they assist in composition by facilitating the expression of ideas. In the form of conceptual structures they give backbone to knowledge representation systems. Within the discipline of library and information science, CVs are used primarily to assist in the retrieval of information. Classifications, thesauri, and subject heading lists are examples of CVs. If vocabulary is widely interpreted to include proper names, then name authority files are also examples of CVs in this domain. The design of CVs within library and information science ranges from activities as simple as listing the allowable values for attributes in a specialized database to tasks as complicated as constructing an in-depth classification scheme. This article looks at the general principles underlying the design of CVs that provide alphabetic and classified approaches to subject indication. It begins with a statement of the purpose of CVs, then discusses the design of their vocabulary, syntax, and semantics. It concludes with a brief review of two issues of present concern: compatibility among CVs and effectiveness of CVs.