On Characterising and Identifying Mismatches in Scientific Workflows
Workflows are gaining importance as a means for modelling and enacting in silico scientific experiments. A major issue which arises when aggregating a collection of analysis operations within a workflow is the compatibility of their inputs and outputs: the analysis operations are supplied by independently developed web services which are likely to have incompatible inputs and outputs. We use the term mismatch to refer to such incompatibility. This paper characterises the mismatches a scientific workflow may suffer from and specifies mappings for their resolution.
KeywordsData Link Analysis Operation Connected Parameter Type Mismatch Representation Mismatch
Unable to display preview. Download preview PDF.
- 1.Belhajjame, K., Embury, S.M., Fan, H., Goble, C.A., Hermjakob, H., Hubbard, S.J., Jones, D., Jones, P., Martin, N., Oliver, S., Orengo, C., Paton, N.W., Poulovassilis, A., Siepen, J., Stevens, R., Taylor, C., Vinod, N., Zamboulis, L., Zhu, W.: Proteome data integration: Characteristics and challenges. In: UK All Hands Meeting (2005)Google Scholar
- 5.Hull, D., Stevens, R., Lord, P., Wroe, C., Goble, C.: Treating shimantic web syndrome with ontologies. In: First Advanced Knowledge Technologies workshop on Semantic Web Services (AKT-SWS 2004) (2004)Google Scholar