Search academic texts

Information × Registration Number 0125U002728, R & D request Title Applying large language models to natural language processing Head Hlybovets Mykola M., Доктор фізико-математичних наук Registration Date 16-06-2025 Organization National University of Kyiv-Mohyla Academy popup.description1 The goal of the project is to create tools for processing, analyzing and generating natural language. Solving NLP problems can potentially erase the barrier in human-computer interaction, because for humans the main means of communication is speech, while computer systems mainly perceive information according to strict protocols. Previously, recurrent and convolutional neural networks were considered the most effective for tasks such as machine translation and language modeling. Now there is a need to research and develop a new architecture of deep neural networks - Transformer, which is extremely effective for solving NLP tasks. Large language models are used for data mining, material discovery and autonomous research, in particular in text processing and analysis - extracting important data from large volumes of text, for example, in legal or medical documents, to accelerate research and decision-making, as well as in search engines to improve the relevance of results by better understanding user queries. Large language models can play a major role in the so-called sentiment analysis of messages on social networks, including for detecting cyber threats. popup.nrat_date 2025-06-16 Close