Web Service Recommendation Based on Latent Features

摘要

Service-oriented computing has been widely recognized as a revolutionary paradigm for software development. Web services have acquired enormous popularity among software developers. This popularity has motivated developers to publish a large number of Web service descriptions in UDDI registries. According to the Web Service Architecture (WSA) published by W3C, users can find services through the repository Universal Description, Discovery and Integration (UDDI). However, the UDDI may find many web services with similar functions, and users may have difficulty deciding which service is suitable. This paper proposes a unified collaborative filtering model for web service recommendation to help users to select the suitable web service. The method in this paper combined the latent and external features of users and services for accurate recommendation. The main advantages of this technique over standard memory-based methods are the higher accuracy, constant time prediction, and an explicit and compact model representation. The experimental evaluation shows that substantial improvements in accuracy over existing methods can be obtained.