A Dynamic QoS-Aware Semantic Web Service Composition Algorithm

The aim of this work is to present a dynamic QoS-aware semantic web service composition algorithm that nds the minimal solution graph that satises the composition request considering multiple QoS criteria and semantic input-output message structure matching restrictions. Our proposal starts computing an initial solution by selecting only those services from the dataset that are relevant to the user request and meet the semantic restrictions. Then, an optimal QoS-aware composition is calculated using Dijkstra shortest path algorithm. Once the solution is obtained, the number of services is minimized using the optimal aggregated QoS value calculated in the previous step as a bound to prune the state space search. Moreover, a set of extensive experiments with ve dierent datasets from the Web Service Challenge 2009-2010 is presented to prove the eciency of our proposal.

keywords: Automatic composition, Shortest Path, QoS optimization, Semantic Web Services