BGCL: Bi-subgraph network based on graph contrastive learning for cold-start QoS prediction

With the advent of Web service technologies, the number of services published on the cloud is increasing rapidly. The quality of service (QoS) becomes a crucial criterion for selecting services from a massive pool of candidates. Collaborative filtering (CF) has become a major way for personalized QoS prediction by leveraging historical interactions between users and services. Due to the increasing number of users and services, CF-based QoS prediction often suffers from data sparsity and coldstart difficulties. Inspired by the advantages of graph contrastive learning in cold-start predictions, we propose BGCL, a bi-subgraph network based on graph contrastive learning to solve the above problems. Firstly, we generate different perspectives of user-neighborhood and service-neighborhood sub-graphs based on sparse user-service bipartite graphs. Next, our model learns user and service embeddings using the graph contrastive learning and graph attention aggregation mechanisms on the generated sub-graphs. Finally, user and service embeddings are fed into a multi-layer perception to predict QoS values. Experimental results show that our model outperforms several existing models in terms of prediction accuracy.

Automated summary

With the increase in the number of services published on the cloud due to Web service technologies, the importance of quality of service (QoS) as a criterion for selection becomes crucial. Collaborative filtering (CF) is a popular approach for personalized QoS prediction, but it faces challenges like data sparsity and cold-start difficulties. To address these issues, the proposed BGCL model leverages graph contrastive learning and attention aggregation mechanisms to generate user and service embeddings and predict QoS values. Experimental results demonstrate that the BGCL model outperforms existing models in terms of prediction accuracy.