Exploiting Web service geographical neighborhood for collaborative QoS prediction

摘要: Abstract Since Web services with equivalent functionalities but different quality are becoming increasingly available on the Internet, predicting unknown QoS value of a service to an active user who has not accessed previously is often required for recommendation and composition. Existing collaborative filtering methods suffer from unavoidable sparsity cold-start problems underestimate role geographical information that inherently exists in user–service rating oriented model. The principal motivation using prediction stems observation ratings perform influenced significantly by their neighborhood, fact verified our empirical data analysis real-world dataset WSDream. Hence, it will be interest incorporate this implicit source prediction. In paper, carefully selected neighbors, clustered bottom-up hierarchical neighborhood clustering method, smoothly integrated into matrix factorization model, thereby building more accurate Further accuracy improvements achieved considering biases users services. experiments WSDream dataset, proposed method outperforms other competitive respect alleviates issues.