Defining Concepts, Generating Knowledge: A Structured Review of Terminological Definitions

Terminological definitions play a central role in the representation, delimitation, and communication of specialized concepts. While they have traditionally been studied within terminology theory, specialized lexicography, and knowledge organization, they have recently acquired renewed relevance in computational settings, including automatic definition extraction, definition generation, ontology learning, and terminology-augmented generation. This paper presents a structured review of research on terminological definitions, tracing the transition from theoretical and methodological approaches to computational and generative perspectives. The review is based on a year-by-year screening of Google Scholar results from 2000 to 2026, complemented by historically relevant works from the 1990s. The paper identifies major research directions, including normative principles for definition writing, corpus-based and communicative approaches, knowledge-oriented modeling, and automatic definition extraction and generation. The review highlights how terminological definitions have progressively moved from manually authored components of terminological resources to objects of computational modeling, evaluation, and reuse.