Support for seamless data exchanges between web services through information mapping analysis using kernel methods

A challenging issue to web services interoperability is seamless data exchanges between web services to be composed. A solution to this problem is to establish semantic mappings from an information item to another. To do that, we present an approximate information mapping analysis. We propose a kernel-based structural similarity measure for XML documents. Simulation results with industrial XML data show that the proposed kernel-based measure outperforms other existing methods.