eLex 2025: Intelligent Lexicography Between Automation and Linguistic Responsibility

Abstract

This paper provides an overview of the international conference Electronic Lexicography in the 21st Century (eLex 2025), whose main theme was Intelligent Lexicography. The conference was held from November 18 to 20, 2025, in Bled, Slovenia, and was dedicated to contemporary trends in electronic lexicography, with particular attention to the application of artificial intelligence and large language models in the creation, enrichment, and evaluation of lexicographic resources. The paper examines the dominant research trends, methodological challenges, and open questions that marked the conference, as well as the positioning of Serbian lexicography in that international context. A special section is devoted to the contributions of researchers from Serbia and their experiences in working on the Dictionary of Contemporary Serbian Language and in experiments with large language models in the field of phraseology and multiword lexical units.

Keywords: electronic lexicography, intelligent lexicography, digital humanities, large language models, multiword lexical units, eLex 2025, Serbian language