Online Program

Applying Business Intelligence Technology to Statistical Processing and Analysis
*Ron J McKenzie, Statistics New Zealand 
Andrew Skerrett, Statistics New Zealand 


Keywords: Business Intelligence, systems, analysis, administrative data, processing

Statistics NZ has successfully applied the emerging technology suite known as Business Intelligence to statistical processing and analysis.

The system combines registers, administrative data series and surveys in a single cell based data warehouse. In a unique approach, raw data is receipted into this store prior to repair. Statistical processing and modelling is driven by workflow transfer of this data into a scenario area. Workflow brokerstransfers of data between the scenario store and the calculation engine. Cube technology is used to examine and compare scenarios. Approved results are written back to the warehouse. A cube farm is deployed to traverse an array of analysis and output needs.

The goal of the system is to provide a framework to free anaslysts from the shackles of the IT development cycle. The framework allows expert end users to define the variable metadata and create the SAS based business rules that are applied by the workflows. The system provides user configured access to Selective Editing and BANFF among other methods. The combination of scenario store and a “whole of dataset” philosophy promises more efficient, less labour intensive editing.