Tag Completion and Refinement for Web Service via Low-Rank Matrix Completion

Abstract

With the development of cloud computing, more and more web services are deployed on the cloud platform. It provides more solutions to the customers, but accompanies with a critical and fundamental problem, that is, how to easily find the desired web services. Since tags provide meaningful descriptions for web services function and non-function properties, some researchers have employed tags to facilitate web service discovery. However, the existing web service tags are often imprecise and incomplete. To complete the missing tags and correct the noisy ones, an efficient web service Tag Completion and Refinement based on Matrix Completion (TagCRMC) approach is proposed. The TagCRMC approach not only models the low-rank property of service-tag matrix, but also simultaneously integrates the content correlation consistency and the tag correlation consistency to ensure the correct correspondence between web services and tags. Experimental results on the real-world web services collection show the encouraging performance of the TagCRMC approach.