Published May 7, 2026 | Version v1

O'ZBEK TILIDA MATNLARNING SENTIMENTAL TAHLILI: ROBERTA VA T5 MODELLARINI SOLISHTIRISH

  • 1. Aniq va Ijtimoiy fanlar universiteti
  • 2. Romanovskiy nomidagi matematika institute

Description

Mazmuniy tahlil (sentiment analysis) — tabiiy tillarni qayta ishlash (NLP) sohasining eng faol rivojlanayotgan yo'nalishlaridan biri bo'lib, u matnlarda ifodalangan his-tuyg'u va munosabatlarni avtomatik aniqlashga xizmat qiladi. Hozirgi kunda ingliz va boshqa yirik tillar uchun yuqori natijaga erishgan transformator arxitekturali modellar o'zbek tili kabi resurslar cheklangan tillar uchun hali etarli darajada o'rganilmagan. Ushbu maqolada o'zbek tilidagi 29 194 ta yozuvdan iborat annotatsiyalangan dataset asosida ikki xil chuqur o'rganish modeli — RoBERTa (BERT asosidagi ketma-ket klassifikator) va UzT5 (T5 asosidagi generativ model) — solishtirilib tekshiriladi. Tajriba natijalari shuni ko'rsatadiki, RoBERTa arxitekturasi asosidagi model 76,5% aniqlik (accuracy) va 75,7% F1-ko'rsatkichi bilan UzT5 modelidan (52,7% aniqlik, 52% F1) sezilarli darajada ustun keladi. Olingan natijalar o'zbek tili uchun diskriminativ klassifikatsiya yondashuvining generativ yondashuvga nisbatan samaradorligini tasdiqlaydi.

Files

V4I228.pdf

Files (443.6 kB)

Name Size
md5:0ed034d1be04aae324ce36416f71a02f
443.6 kB Preview Download

Additional details

References

  • Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., … & Liu, P. J. (2020). Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer. JMLR, 21(140), 1-67.
  • Brown, T., et al. (2020). Language Models are Few-Shot Learners. Advances in Neural Information Processing Systems, 33, 1877-1901.
  • Mansurov, B., Mansurov, A., & Ikhtiyor, R. (2021). UzBERT: O'zbek tili uchun oldindan o'qitilgan til modeli. arXiv preprint arXiv:2110.12693.
  • Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., … & Polosukhin, I. (2017). Attention is All You Need. Advances in Neural Information Processing Systems, 30.
  • Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., … & Stoyanov, V. (2019). RoBERTa: A Robustly Optimized BERT Pretraining Approach. arXiv preprint arXiv:1907.11692.
  • Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. NAACL-HLT 2019, 4171-4186.
  • Liu, B. (2012). Sentiment Analysis and Opinion Mining. Synthesis Lectures on Human Language Technologies, 5(1), 1-167.
  • Abdurakhmonova, N., & Matlatipov, G. (2022). O'zbek tilini kompyuter asosida qayta ishlash: muammolar va istiqbollar. O'zbekiston Milliy universiteti xabarlari.