IDENTIFICATION OF KNOWN AND EMERGENT EUPHEMISTIC EXPRESSIONS IN ENGLISH CORPORA: A COMPARATIVE ANALYSIS OF CORPUS-BASED APPROACHES

Abstract

This thesis examines the difference between classifying a known potentially euphemistic expression in context and discovering an expression that is absent from an existing lexicon. A focused review of published computational studies shows why the two tasks require different corpus sampling and evaluation procedures. A staged workflow is proposed for collecting candidate expressions, checking their local meanings, and assessing whether a method can recover previously unseen forms. The workflow is a research proposal; no new corpus or performance result is reported.