Lexicographer and consultant, Washington, DC
Pages: 6 Published: Jan 1995
Term ambiguity primarily as the result of interdisciplinary borrowing has created problems for the unambiguous retrieval of information from textbases using terms as search tools. The addition of qualifiers to terms facilitates more precise information retrieval, but human editing to install such qualifiers in texts is time-consuming. The paper explores means for analyzing and automating the inclusion of qualifiers in order to improve retrieval. Examples from the NASA SCIENTIFIC AND TECHNICAL INFORMATION database serve to illustrate options for automating the identification of terms having specific conceptual meanings. The article addresses issues of Boolean searches, the use of controlled vocabulary, and frequency ranking, as well as the difference between ‘words’ in general language and ‘terms’ in technical and special languages, underscoring the value of disambiguated terms as a tool for information retrieval.
term ambiguity, information retrieval, textbases, Boolean algebra, apparent ambiguity, manual editing, automated editing, controlled vocabulary, term meaning
Paper ID: STP13747S