One size does not fit all: logic-based clustering for on-the-fly web service composition and verification

Recently, web service composition (WSC) has been widely emerging since it is obviously hopeless to develop a specific web service which can singlehandedly fulfil completely a requirement posed by clients. Moreover, as a WSC solution often needs to satisfy various kinds of constraints, its correctness is also required to be formally verified. However, composition and verification tasks always suffer from huge computational cost when applied in practice. It prompts the consideration of clustering approaches. Unfortunately, the existing web service clustering approaches accomplish neither the soundness nor completeness of a composition solution. To overcome this problem, we propose a logic-based approach for clustering web services, where logic formulae are used for representing web services and their clusters. The clustering results are then further applied for WSC in an on-the-fly manner. Our approach achieves both soundness and completeness. In experimentation, our approach also improves the performance of WSC process significantly.