OSGNN: Original graph and Subgraph aggregated Graph Neural Network

Yan, Yeyu; Li, Chao; Yu, Yanwei; Li, Xiangju; Zhao, Zhongying

doi:10.1016/j.eswa.2023.120115

Cited by 6 publications

(1 citation statement)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Neural Network (GNN) is a general term for algorithms that utilize neural networks to learn graph data, extract features from graph-structured data, discover patterns, and perform graph-learning tasks[14]. Graph neural networks transform graph-structured data into canonical and standardized representations and input them into many different neural networks for training, achieving excellent results in tasks such as node classification, edge information propagation, and graph clustering.…”

mentioning

confidence: 99%

A personalized recommendation algorithm for online teaching resources of vocal music based on graph neural network

Wang

2023

Applied Mathematics and Nonlinear Sciences

View full text Add to dashboard Cite

In the development process of education informatization, realizing personalized learning is the urgent demand of learners. In this paper, we construct an online teaching resources recommendation algorithm (VGNN) based on a graph neural network (GNN). The project-based collaborative filtering algorithm is used to find highly similar teaching resource materials using Pearson’s correlation coefficient in order to personalize the recommendation of online teaching resources. The personalized recommendation algorithm for online teaching resources was tested and analyzed using the online teaching course Fundamentals of Vocal Music Theory as a case study. The results show that the course “Fundamentals of Vocal Music Theory” is divided into 4 clusters with clear boundary contours under the graph neural network algorithm, and the number of knowledge points in the 4 clusters is 2308, 655, 513, and 97, respectively. After learning the personalized recommended content from the 4 clusters of knowledge points, the average rate of loss of points for each of them decreases by about 0.3. The algorithm effectively solves the accurate classification of vocal knowledge and implements personalized recommendations for educational resources.

show abstract

mentioning

confidence: 99%