Time-aware service recommendation for mashup creation in an evolving service ecosystem

Abstract

Web service recommendation has become increasingly important as services become increasingly prevalent on the Internet. Existing methods either focus on content matching techniques such as keyword search and semantic matching, or rely on Quality of Service (QoS) prediction. However, the fact that services and their mashups typically evolve over time has not been given sufficient attention. We argue that a practical service recommendation approach should take into account the evolution of services in the context of a service ecosystem. In this paper, we present a method to extract service evolution patterns by exploiting Latent Dirichlet Allocation (LDA) and time series prediction. Based on it, we have developed a time-aware service recommendation framework guiding mashup creation seamlessly integrating service evolution, collaborative filtering and content matching. Experiments on real-world ProgrammableWeb data set show that our approach leads to a higher precision than traditional collaborative filtering and content matching methods.