Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 10, 2025Journal of Computational Methods in Sciences and Engineering

Application of graph neural networks (GNNs) in learning situation prediction and educational resource recommendation algorithms

View Full Paper
Ask AI
Bookmark
Share

Authors

XLXuan Liu

Discussion

Loading...

Member takes

Overview

This observational analysis enhances educational recommendation and dropout risk prediction, suggesting effective learning behavior modeling in diverse contexts.

Key Points

  • The proposed HetGNN model significantly improves dropout risk prediction, achieving an F1-score of 0.91.
  • The relational graph convolutional network effectively aggregates student-resource-knowledge relationships, ensuring enhanced educational outcomes.
  • Dynamic attention generated by the Transformer encoder captures evolving engagement, leading to better learning status forecasting.
  • Improvements in resource recommendation show a notable NDCG of 0.93, addressing previously underrepresented long-tail resources.

Cite This Study

Xuan Liu (2025) studied this question.

synapsesocial.com/papers/68c1dda254b1d3bfb60fc4a4https://doi.org/10.1177/14727978251374326
View Full Paper
Ask AI
Bookmark
Share