AI-Enhanced Transdisciplinary Data Encoding for LLMs Training

Makhachashvili, Rusudan та Bober, Nataliia (2025) AI-Enhanced Transdisciplinary Data Encoding for LLMs Training Proceedings of the 29th World Multi-Conference on Systemics, Cybernetics and Informatics: WMSCI 2025, 1. с. 327-333. ISSN 2771-0947

Текст SA747NT.pdf Restricted to Тільки персонал архіва Download (383kB) | Запит копії

Анотація

The rapid advancement of artificial intelligence (AI) has reshaped linguistic data encoding, particularly for Large Language Models (LLMs). AI-driven annotation techniques enable efficient lexical processing, semantic disambiguation, and automated neology tagging, refining computational language modeling across transdisciplinary domains. This study explores AI-enhanced methodologies for encoding linguistic data for LLM training. AI-assisted lexicographic workflows enable LLMs to dynamically adjust to linguistic evolution while ensuring scalable annotation across diverse transdisciplinary corpora. LLMs trained on transdisciplinary lexicons can generate cross-modal language interpretations, refining machine-generated discourse across domains. The inquiry objective is the investigation of the innovative philosophic aspects cyberspace through the lenses of the language development processes as it informs AI models elaboration, LLMs training, and digital communication. The study design is the disclosure of cyberspace as an ontology model and as a logosphere model. Two data encoding projects, developed by the authors, serve as foundational elements for this investigation. A methodology and AI-augmented, AI-performed protocols of computer vocabulary innovative elements phenomenological features identification is introduced supplying the template for a new study field – phenomenological, AI-enhanced digital neology, neography and neosemiotics. Transdisciplinary educational applications of these approaches to data encoding, include: training AI-enhanced NLP models for transdisciplinary communication; developing standardized linguistic annotation protocols, ensuring interoperability across AI-driven lexicographic systems; integrating transdisciplinary discourse structures into machine-learning lexicons, refining AI adaptive language comprehension.

Тип елементу :	Стаття
Ключові слова:	large language models; generative artificial intelligence; artificial intelligence linguistics; linguistic philosophy; innovative logosphere of cyberspace; language data encoding; digital communication
Типологія:	Статті у базах даних > Scopus (без квартилю)
Підрозділи:	НДЛ інтернаціоналізації вищої освіти
Користувач, що депонує:	Русудан Кирилевна Махачашвілі
Дата внесення:	10 Груд 2025 13:55
Останні зміни:	10 Груд 2025 13:55
URI:	https://elibrary.kubg.edu.ua/id/eprint/54698

Actions (login required)

Перегляд елементу