Raqamli Transformatsiya va Sun’iy Intellekt
AKADEMIK MAQOLALARDAN METAMA’LUMOTLARNI EKSTRAKSIYA QILISHDA MAVJUD YECHIMLAR
1 Qo‘qon davlat universiteti
Keywords
BERT
katta til modellari
multimodal yondashuv
akademik maqolalar
GPT-4
metama’lumotlarni ekstraksiya qilish
Abstract
Ushbu maqolada katta til modellarining akademik maqolalardan metama’lumotlarni ekstraksiya qilishdagi imkoniyatlari tahlil qilinadi. GPT-4, BERT va LLaMA kabi modellar an’anaviy usullarga nisbatan yuqori moslashuvchanlik, murakkab ma’lumotlarni ajratish va ko‘p tilli ishlash afzalliklariga ega. ChatExtract usuli 90% F1-natijaga erishgan. Biroq resurs talabchanligi, faktik xatolar va maxfiylik muammolari asosiy cheklovlardir. Tor sohaga moslashtirilgan modellar tejamkorligi bilan ajralib turadi.
References
- Alyafeai, Z., Al-Shaibani, M. S., & Ghanem, B. (2025). MOLE: Metadata extraction and validation in scientific papers using LLMs. arXiv preprint arXiv:2505.19800.
- Usta, G. (2026). Automating MARC Records from Turkish Title Pages: A Performance and Efficiency Comparison of Fine-Tuned BERT vs. VLMs. Journal of Library Metadata, 1–19. https://doi.org/10.1080/19386389.2026.2627732
- Kohút, J., Dočekal, M., Hradiš, M., Vaško, M. (2026). BiblioPage: A Dataset of Scanned Title Pages for Bibliographic Metadata Extraction. In: Yin, XC., Karatzas, D., Lopresti, D. (eds) Document Analysis and Recognition – ICDAR 2025. ICDAR 2025. Lecture Notes in Computer Science, vol 16025. Springer, Cham. https://doi.org/10.1007/978-3-032-04624-6_17
- J. Dagdelen et al., “Structured information extraction from scientific text with large language models,” Nature Communications, vol. 15, no. 1, 2024, doi: 10.1038/s41467–024-45563-x.
- Johan Krause, Igor Shapiro, Tarek Saier, and Michael Färber. 2021. Bootstrapping Multilingual Metadata Extraction: A Showcase in Cyrillic. In Proceedings of the Second Workshop on Scholarly Document Processing, pages 66–72, Online. Association for Computational Linguistics.
- T. Saier, M. Ohta, T. Asakura, and M. Färber, “HyperPIE: Hyperparameter Information Extraction from Scientific Publications,” Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14609 LNCS, pp. 254–269, 2024, doi: 10.1007/978-3-031-560606_17.
- Polak, M.P., Morgan, D.: Extracting accurate materials data from research papers with conversational language models and prompt engineering - Example of ChatGPT (Mar 2023). https://doi.org/10.48550/arXiv.2303.05352
- Boukhers, Z., & Yang, C. (2025). Comparison of Feature Learning Methods for Metadata Extraction from PDF Scholarly Documents. arXiv preprint arXiv:2501.05082.
- Choi, W., Yoon, H. M., Hyun, M. H., Lee, H. J., Seol, J. W., Lee, K. D., ... & Kong, H. (2023). Building an annotated corpus for automatic metadata extraction from multilingual journal article references. PloS one, 18(1), e0280637.