GRAF NEYRON TARMOQLARI YORDAMIDA AXBOROT QIDIRUV TIZIMLARINI TAKOMILLASHTIRISH

Shukurboyeva S. F.


Annotatsiya

Zamonaviy axborot qidiruv tizimlari bugungi kunda foydalanuvchi so‘rovlariga aniq va relevant takliflarni taqdim etishda bir qator muammolarga duch kelmoqda. Tadqiqotda graf neyron tarmoqlari (GNN) yordamida hujjatlar o‘rtasidagi murakkab munosabatlarni modellashtirish va qidiruv tizimining samaradorligini oshirishning yangi yo‘llari izlandi. WikiIR hamda MS MARCO kabi yirik ma’lumotlar to‘plamlarida o‘tkazilgan eksperimentlar, an’anaviy BM25 metodiga qaraganda MAP ko‘rsatkichining 18.3% ga, NDCG@10 natijalarining esa 22.7% ga yaxshilanganini ko‘rsatdi. Olingan natijalar GNN modellarining semantik qidiruv vazifalarida istiqbolliligini tasdiqlaydi.


Adabiyotlar

  1. Devlin, J., et al. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. NAACL-HLT.
  2. Kipf, T. N., & Welling, M. (2017). Semi-Supervised Classification with Graph Convolutional Networks. ICLR.
  3. Velickovic, P., et al. (2018). Graph Attention Networks. ICLR.
  4. Liu, Z., et al. (2020). Graph Neural Networks for Information Retrieval. SIGIR.
  5. Nogueira, R., & Cho, K. (2019). Passage Re-ranking with BERT. arXiv preprint.
  6. Hamilton, W., et al. (2017). Inductive Representation Learning on Large Graphs. NeurIPS.
  7. Yoon, S., et al. (2020). Graph Neural Networks for Text Classification: A Survey. arXiv preprint.
  8. Wang, S., et al. (2020).Neural Models for Information Retrieval. Foundations and Trends® in Information Retrieval.
  9. Wu, Z., et al. (2021).A Comprehensive Survey on Graph Neural Networks. IEEE Transactions on Neural Networks and Learning Systems.
  10. Ying, R., et al. (2018).Graph Convolutional Neural Networks for Web-Scale Recommender Systems. KDD.