Web service schema matching based on invocation instance duplicates

Abstract

Web service schema matching is the preliminary phase of service composition and service Mashup. With the entry of large numbers of services on the Internet, it will be difficult to find and suggest the most relevant service candidates for new composition and Mashup. Furthermore, the composition and Mashup candidates must be confirmed by manual work which is very inefficient and even inefficacious. In this work, we present an approach for web service schema matching based on service invocation instance duplicates. This approach predicts the schema relationship between web service parameters based on duplicates among invocation data sets and is very different from the syntactical-based and semantic-based matching approaches. It can be used for predicting and verifying service composition and Mashup candidates.