BR21882268 " Automated Construction of a Multilingual Ontology for Empowering the Kazakh Language through Advanced AI Technologies"

Relevance

The developed methods and algorithms have a sufficient degree of efficiency and have been tested at the level of the conducted experiments.

Purpose

The purpose of this work is to develop tools for automated content filling of an intelligent NLP information resource, which implies the creation and implementation of effective algorithms and technologies capable of processing and analyzing large amounts of text data in different languages to extract useful information and knowledge for a given subject area.

Expected and achieved results

Significant results have been achieved within the framework of the project during the reporting period. Methods of automatic extraction of entity names for ontology from natural language texts have been developed, and their experimental study has been conducted on scientific publications on NLP. The terminological core of NLP ontology has been updated, which now includes descriptions of methods, their implementations, pre-trained models, and other information resources. A specialized data warehouse has been created for the NLP intellectual information resource and tools have been developed for automated filling of its content. In addition, a web-based user interface was created, and the quality of the resource was evaluated by an expert group. An important result was also the publication of articles in highly rated scientific journals included in the Web of Science and Scopus databases, as well as in publications recommended by the COKNVO. All the tasks of the project comply with modern scientific standards and have a high degree of novelty.

Research team members with their identifiers (Scopus Author ID, Researcher ID, ORCID, if available) and links to relevant profiles

1 Scientific supervisor Musabaev Rustam Rafikovich (rmusab@gmail.com )

2 Responsible executor Toleu Alymzhan alymzhan.toleu@gmail.com

3 Performer Barakhnin Vladimir Borisovich barakhninvb@yandex.ru

4 Performer Krasovitsky Alexander Mikhailovich akrassovitskiy@gmail.com

5 Performer Tolegen Gulmira gulmira.tolegen.cs@gmail.com

6 Performer Ravil Rafikovich Musabaev ravmus@gmail.com

7 Performer Kozbagarov Olzhas Borlykovich kozbagarov@yahoo.com

List of publications with links to them

1. Mussabayev R., Mussabayev R. High-Performance Hybrid Algorithm for Minimum Sum-of-Squares Clustering of Infinitely Tall Data //Mathematics. – 2024. – Т. 12. – №. 13. – С. 1930.
2. Kozbagarov O., Mussabayev R. Distributed random swap: An efficient algorithm for minimum sum-of-squares clustering //Information Sciences. – 2024. – Т. 681. – С. 121204.
3. Baktibayev D. et al. Literature review on aftershock and earthquake prediction models aided by NLP summarization and ontology extraction techniques //Procedia Computer Science. – 2024. – Т. 238. – С. 579-586.
4. Toleu A., Tolegen G., Mussabayev R. Topic Modeling with Variable Neighborhood Search //International Conference on Computational Collective Intelligence. – Cham : Springer Nature Switzerland, 2024. – С. 234-246.
5. Kozbagarov O. et al. Interpretable Dense Embedding for Large-Scale Textual Data via Fast Fuzzy Clustering //International Conference on Computational Collective Intelligence. – Cham : Springer Nature Switzerland, 2024. – С. 206-218.
6. Tolegen G., Toleu A., Mussabayev R. Enhancing Low-Resource NER via Knowledge Transfer from LLM //International Conference on Computational Collective Intelligence. – Cham : Springer Nature Switzerland, 2024. – С. 238-248.

Mussabayev R., Mussabayev R. Superior parallel big data clustering through competitive stochastic sample size optimization in big-means //Asian Conference on Intelligent Information and Database Systems. – Singapore : Springer Nature Singapore, 2024. – С. 224-236.