Automating Scientific Experiments on the Semantic Grid

Abstract

We present a framework to facilitate automated synthesis of scientific experiment workfiows in Semantic Grids based on high-level goal specification. Our framework has two main features which distinguish it from other work in this area. First, we propose a dynamic and adaptive mechanism for automating the construction of experiment workflows. Second, we distinguish between different levels of abstraction of loosely coupled experiment workflows to facilitate reuse and sharing of experiments. We illustrate our framework using a real world application scenario in the physics domain involving the detection of gravitational waves from astrophysical sources. © Springer-Verlag 2004.