Title: Construction of an online vocal teaching model based on node influence and knowledge graph

Authors: Ming Tian

Addresses: School of Music, Xi'an University, Xian, Shaanxi, China

Abstract: In response to the cold start and data sparsity issues faced by online teaching resource recommendation methods, this study first proposes using a node influence model to measure the influence of entities in the knowledge graph. Then, a vocal teaching resource recommendation model and learning effectiveness evaluation model based on long short-term memory network are established. The results show that the recommendation accuracy of the proposed model is 68.23%, and the area under the curve is 73.76%, which is 5.06% and 7.34% higher than the traditional RippleNet model, respectively. The accuracy and recall of the proposed learning effect prediction model are 0.525 and 0.224, respectively, which are 87.43% and 25.45% higher than traditional long short-term memory networks. The experimental results have demonstrated the recommendation and evaluation performance of the proposed model, which helps to improve teaching effectiveness and promote the development of online vocal teaching.

Keywords: node influence; vocal teaching; knowledge graph; resource recommendation.

DOI: 10.1504/IJWMC.2026.154163

International Journal of Wireless and Mobile Computing, 2026 Vol.30 No.4, pp.368 - 381

Received: 28 Oct 2024
Accepted: 24 Apr 2025
Published online: 15 Jun 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article