作者: Jihie Kim , Yolanda Gil , Varun Ratnakar
DOI: 10.1007/11926078_26
关键词: Metadata repository 、 Data element 、 Computer science 、 World Wide Web 、 Workflow 、 Meta Data Services 、 Metadata modeling 、 Executable 、 Semantic Web 、 Metadata
摘要: In recent years, workflows have been increasingly used in scientific applications. This paper presents novel metadata reasoning capabilities that we developed to support the creation of large workflows. They include 1) use semantic web technologies handling constraints on file collections and nested collections, 2) propagation validation from inputs outputs a workflow component, through links among components workflow, 3) sub-workflows generate needed for creation. We show how these executable an earthquake science application with more than 7,000 jobs, generating 100,000 new files.