Semantic type checking in scientific workflows

Abstract

Scientists are increasingly utilizing Grids to manage large data sets and execute scientific experiments on distributed resources [1]. Scientific workflows are used as means for modelling and enacting scientific experiments [2]. Windows Workflow Foundation (WF) is a major component of Microsoft’s .NET technology which offers lightweight support for long-running workflows. It provides a comfortable graphical and programmatic environment for the development of extended BPEL-style workflows but offers little support for ensuring that the resulting workflows are complete, robust and meaningful in the user’s scientific domain.

Workflow building tools rely on the developer’s understanding of multiple services and the data required to execute them. Syntactic type definitions of these data are not meaningful enough to ensure type safety, which are only discovered during execution. We aim to enrich type definitions with semantics in order to guide developers to resolve type mismatch issues at design time.

The approach we have taken is to develop SAWDL-compliant annotations for workflow and use them with a semantic reasoned to guarantee semantic type correctness in scientific workflows.

More information

Identifiers

Catalogue record

Export record