Animated WUN Logo Grid Logo
Short WUN title

Demonstrator: Seamless Access to Multiple Data sets

Ease of access to multiple data sets is becoming increasingly important to researchers, and new techniques and methods to facilitate this are being developed. In addition to holding different types of information, these databases/data sets may be geographically dispersed.

The demonstrator problem chosen for SAMD examines the asymmetric effects of interest rate changes on the UK's Gross Domestic Product (GDP). This is a substantive research question of genuine policy interest, based on data drawn from the National Statistics Databank hosted at the University of Manchester. It contains some 55,000 time series in 34 major data sets from the UK Office for National Statistics relating to economics, trade, employment and industry. A subset of time series related to interest rates and GDP is selected via the SAMD user interface, and then sent to an HPC engine on the Grid for analysis.

An array of first difference of the logarithm of real GDP is compared with a series of arrays of earlier changes in interest rates. The results show that changes in interest rates have greater effect on output when past growth has been high than when past growth has been normal or negative, in other words, the effect of interest rate changes on GDP varies over the business cycle.

SAMD demonstrates the successful incorporation of emerging Grid technologies into an existing social science data service. It shows how the integration of access to both data and computational resources within a single sign-on environment enables the automation of complex workflows, facilitating the scaling up of social science research applications. Finally, it shows how adding programmable interfaces (protocols) to existing services facilitates the development of third-party, value-adding client applications.

Valid XHTML 1.0! is maintained by David De Roure
Last updated 24 January 2005.