Published May 6, 2026 | Version v1
Journal article Open

GETEROGEN GRAF NEYRON TARMOG'I ASOSIDA O'QUV REJA OPTIMALLASHTIRISH VA SHAXSIY TAVSIYALAR TIZIMI: KOMPYUTER INJINIRINGI MUTAXASSISLIGI MISOLIDA

Authors/Creators

  • 1. Urganch davlat Univesiteti

Description

Ushbu maqolada o'quv reja optimallashtirish va shaxsiy akademik tavsiyalar uchun Geterogen Graf Neyron Tarmog'i (GGNT) arxitekturasi taklif etiladi. Tizim talaba, fan va kategoriya tugunlarini birlashtiruvchi uch qavatli geterogen grafni quradi va ikki qavatli GraphSAGE aggregatsiyasi orqali ko'p darajali yashirin bog'liqliklarni kodlaydi. SHAP asosida tushuntirish moduli har bir tavsiyani asoslab beradi. 184 talaba, 109 fan va 8 486 yozuvdan iborat dataset Kompyuter Injiniringi datasetida o'tkazilgan tajribalar GGNT tizimining NDCG@10 ko'rsatkichi bo'yicha K-Means+Kosinus tayanch usulidan 19,6% ustun ekanligini ko'rsatdi. Random Forest baho bashoratchisi 97,5% aniqlikka erishdi.

Files

V4I225.pdf

Files (455.0 kB)

Name Size
md5:75d1021231ce5f03b0906edecb178e3a
455.0 kB Preview Download

Additional details

References

  • Arrieta A.B. et al. Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities // Information Fusion. — 2020. — Vol. 58. — P. 82–115.
  • Wang H. et al. RippleNet: Propagating user preferences on the knowledge graph // Proc. CIKM 2018. — 2018. — P. 417–426.
  • Lundberg S.M. et al. From local explanations to global understanding with explainable AI for trees // Nature Machine Intelligence. — 2020. — Vol. 2, No. 1. — P. 56–67.
  • Burke R. Hybrid recommender systems: Survey and experiments // User Modeling and User-Adapted Interaction. — 2002. — Vol. 12, No. 4. — P. 331–370.
  • Wang X. et al. Heterogeneous graph attention network // Proc. WWW 2019. — 2019. — P. 2022–2032.
  • Drachsler H., Greller W. Privacy and analytics: It's a DELICATE issue // Proc. LAK'16. — 2016. — P. 89–98.
  • Adomavicius G., Tuzhilin A. Toward the next generation of recommender systems // IEEE Trans. Knowledge and Data Engineering. — 2005. — Vol. 17, No. 6. — P. 734–749.
  • Xoliqov M.A., Mirzayev I.B. Axborot-kommunikatsiya texnologiyalari asosida ta'limni individuallashtirish // O'zbekiston Milliy universiteti xabarlari. — 2023. — №1. — B. 45–52.
  • Lundberg S.M., Lee S.-I. A unified approach to interpreting model predictions // NeurIPS 2017. — 2017. — P. 4765–4774.
  • Hamilton W.L., Ying R., Leskovec J. Inductive representation learning on large graphs // NeurIPS 2017. — 2017. — P. 1024–1034.
  • Kipf T.N., Welling M. Semi-supervised classification with graph convolutional networks // Proc. ICLR 2017. — 2017.
  • He X. et al. LightGCN: Simplifying and powering graph convolution network for recommendation // Proc. SIGIR 2020. — 2020. — P. 639–648.
  • Koren Y., Bell R., Volinsky C. Matrix factorization techniques for recommender systems // Computer. — 2009. — Vol. 42, No. 8. — P. 30–37.
  • Khorasani E.S. et al. Applying recommender systems for learning management systems // Proc. Int. Conf. Web and Open Access to Learning. — 2014.