Hybrid Adaptive Web Service Selection with SAWSDL-MX and WSDL Analyzer

Matthias Klusch; Patrick Kapahnke; Ingo Zinnikus

In: Lora Aroyo; Paolo Traverso; et al. (Hrsg.). The Semantic Web: Research and Applications. European Semantic Web Conference (ESWC-2009), 6th, May 31 - June 4, Heraklion, Greece, Pages 550-564, Lecture Notes in Computer Science (LNCS), Vol. 5554, IOS Press, 2009.


In this paper, we present an adaptive, hybrid semantic matchmaker for SAWSDL services, called SAWSDL-MX2. It determines three kinds of semantic service similarity with a given service request, that are logic-based, text-based and structural similarity. In particular, the degree of structural service similarity is computed by the WSDL-Analyzer tool [12] by means of XMLS tree edit distance measurement, string-based and lexical comparison of the respective XML-based WSDL services. SAWSDL-MX2 then learns the optimal aggregation of these different matching degrees over a subset of a test collection SAWSDL-TC1 based on a binary support vector machine-based classifier. Finally, we compare the retrieval performance of SAWSDL-MX2 with a non-adaptive matchmaker variant SAWSDL-MX1 [1] and the straight forward combination of its logic-based only variant SAWSDL-M0 with WSDL-Analyzer.

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence