Grid-enabling Social Scientists: the FINGRID infrastructure
Lee Gillam, Khurshid Ahmad, Gary Dear · 2004
Abstract. This paper discusses a Grid implementation for facilitating large-scale social sciences research. Grid-enabling resources, the computers, programs and data, necessitates technological planning, even in hard science disciplines like particle physics and astronomy where the resources are well identified. Grid middleware is inherently complex, covering a broad spectrum of functionality, and requires some efforts to understand and use effectively. This paper describes how emergent grid middleware (Globus Toolkit, OGSA-DAI, Condor, SRB) was used to create a 24-node grid for managing, retrieving and analysing “live ” data. The application area was financial economics and econometrics, where time serial and textual data were analysed concurrently: speed-ups of upto 5-10 times were noted in both cases. The Grid is being used for capacity building in both research and teaching.