DOI: https://doi.org/10.36719/2663-4619/127/171-179
Emma Seidova 
Odlar Yurdu University
Baku, Azerbaijan
PhD student
seidova.emma@oyu.edu.az
Javid Babayev
Nakhchivan State University
Nakhchivan, Azerbaijan
PhD in Philology
cavidbabayev@ndu.edu.az
Sayyara Sadikhova
Nakhchivan State University
Nakhchivan, Azerbaijan
PhD in Art Criticism
seyyaresadixova@ndu.edu.az
Gunay Aghayeva
Azerbaijan University of Languages
Baku, Azerbaijan
PhD in Philology
gunay.agayeva@adu.edu.az
Zhala Mammadova
Azerbaijan University of Languages
Baku, Azerbaijan
PhD in Philology
mammadova_jala@adu.edu.az
AI-Powered Lexical Innovation: Modeling Neologisms with NLP
Abstract
The article explores the ways of modeling neologisms through NLP (Natural Language Processing) that is one of the means of forming lexical innovation. The rapid evolution of digital communication has accelerated the creation and diffusion of neologism. They are considered to be newly coined words or expressions that reflect emerging cultural, technological, and social realities. Traditional lexicographic methods often struggle to capture these innovations in real time. This article elaborates how Artificial Intelligence (AI) and NLP can be applied to model, detect, and analyze neologisms in large-scale text corpora. Using an NLP pipeline incorporating word embedding models, contextual language models, and frequency-based detection, the study demonstrates how AI-driven approaches can identify lexical innovations and track their semantic development. AI-powered models can detect patterns in existing vocabulary, infer morphological rules, and even generate plausible novel words that reflect contemporary linguistic usage. By leveraging neural networks, probabilistic models, and contextual embeddings, researchers can simulate processes of word formation, predict the likelihood of adoption, and analyze semantic integration of neologisms into the broader lexicon. Results show that contextual transformer models significantly outperform static embedding techniques in detecting emerging vocabulary. The findings highlight the potential of AI systems to support lexicography, sociolinguistics, and digital humanities by providing scalable methods for monitoring linguistic change.
Keywords: NLP, neologisms, word embedding models, Artificial intelligence