Service mining for Web service composition

In this paper, we propose the concept of service mining and show how this mining helps in the automatic assemblage of services into complex aggregates called composite set-vices. Differences of service mining with Web mining are explained. With a framework for composite service processing as background, possible aspects of service mining are explored. We introduce the concept of service constraint processing (SCP) for service mining, and model it as a 2-tier constraint satisfaction problem (CSP) problem. An algorithm for solving the SCP problem is also given.