Article: An intelligent recommendation system for music therapy resources based on a knowledge graph Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.40 pp.18 - 33 Abstract: To address the limitation of insufficient flexibility in current music therapy recommendation systems for capturing critical inter-entity relationships, this paper first uses the pre-trained language model BERT to enrich representation vectors of entities and relationships, and integrates the user-music therapy resource interaction graph into the knowledge graph, extracting collaborative information. The graph attention mechanism is introduced, allowing the weight of each neighbour node to be dynamically adjusted according to its relationship with the target node. Finally, the score that the user gives to the music therapy resource is predicted by calculating the dot product between the user representation and the music therapy resource representation, and the top N music therapy resources are recommended. Experimental results show that the AUC of the proposed model is improved by 4.85-21% compared to the baseline model, 21%, which can accurately recommend music therapy resources that match the user's preferences. Inderscience Publishers - linking academia, business and industry through research

Title: An intelligent recommendation system for music therapy resources based on a knowledge graph

Authors: Ya Ni; Yingjin Shan

Addresses: School of Humanities and Law, China University of Petroleum (East China), Qingdao, 266580, China ' Academic Affairs Office, China University of Petroleum (East China), Qingdao, 266580, China

Abstract: To address the limitation of insufficient flexibility in current music therapy recommendation systems for capturing critical inter-entity relationships, this paper first uses the pre-trained language model BERT to enrich representation vectors of entities and relationships, and integrates the user-music therapy resource interaction graph into the knowledge graph, extracting collaborative information. The graph attention mechanism is introduced, allowing the weight of each neighbour node to be dynamically adjusted according to its relationship with the target node. Finally, the score that the user gives to the music therapy resource is predicted by calculating the dot product between the user representation and the music therapy resource representation, and the top N music therapy resources are recommended. Experimental results show that the AUC of the proposed model is improved by 4.85-21% compared to the baseline model, 21%, which can accurately recommend music therapy resources that match the user's preferences.

Keywords: music healing resource; intelligent recommender system; BERT model; knowledge graph; graph attention network.

DOI: 10.1504/IJICT.2025.149812

International Journal of Information and Communication Technology, 2025 Vol.26 No.40, pp.18 - 33

Received: 26 Jul 2025
Accepted: 18 Sep 2025
Published online: 13 Nov 2025 *

Title: An intelligent recommendation system for music therapy resources based on a knowledge graph

Keep up-to-date