Riding the open source wave is to combine the best technologies currently available and implementing missing technologies to provide a framework, which strives in building a collaborative community. The emergence of web-based frameworks provides researchers opportunities to access large pool of data for online introspection.

Our open source Web-based application aims to develop an open and a common framework for statistical modelling. In this paper we demonstrate a web-based prototype framework based on Java and R-project. The current implementation features include ability to explore relationships between socio-spatial data (such as voting data of 2007 federal election and 2006 census data) using statistical functions, such as regression analysis, for a user selected area in Australia.

The application framework consists of a server side Java application interfacing R-project with rJava providing a sound statistical analysis package. While being an online application it is also able to provide results offline for further investigation in a PDF format.

The framework employs a bottom-up methodology where the statistical analysis is first implemented in R-project. Next an rJava application is used to collate the statistical information and appropriate analysis results are despatched to a JSP (Java Server Page) web frontend. The framework components R-project and rJava interface allow easy integration with databases and Java/JSP based web frontends. This approach not only decouples the computing complexity at the end user by avoiding any installation and configuration of applications, but also enhances the capacity of researchers to undertake socio-spatial data analysis.