Improving Data Quality for Web Services Composition
Xitong Li, Stuart Madnick, Hongwei Zhu, Yushun Fan · 2009
Independently developed Web services often have different assumptions about the interpretation of the exchanged data, such as inconsistent data representation, unit, precision, and scaling. In practice, data misinterpretation results in many data quality problems and further hampers the execution of service composition. In this paper, we present a context-based mediation approach to handle inconsistent data interpretations and improve data quality for Web services composition. The assumptions about data interpretation of the involved services are made explicit and represented as contexts. A common ontology is defined to describe the contexts. Necessary conversions between elements of the contexts are implemented using XPath functions and external (e.g., third-party) services to reconcile inconsistent contexts. The WSDL descriptions of Web services are annotated with appropriate contexts using the W3C standard SAWSDL. Given a naïve composition ignoring contexts, the reasoning engine can automatically detect context conflicts within the naïve composition and reconcile these conflicts by producing a mediated composition that incorporates appropriate conversions. A proof-of-concept prototype, Context Mediation Tool (CMT), has been developed to validate and demonstrate the approach. 1.