Open DMIX - Data Integration and Exploration Services for Data Grids, Data Web and Knowledge Grid Applications
D. Hanley, Xinwei Hong, G. Bhagavantha Rao · 2003
Although the bulk of the research focus has been on the second phase, most data mining projects spend the majority of the time and effort on the first and third phases. This will certainly continue to be the case as data mining is applied to distributed data using grid based and web based computing platforms. It is convenient to think of the output of the data mining phase as a statistical or data mining model, which can be expressed using the XML Predictive Model Markup Language or PMML [11]. The PMML model can then be used in the deployment phase. From this perspective, the PMML model captures an important part of what is sometimes called the “grid intelligence” of a distributed data grid, data web, or knowledge grid application. In this paper, we introduce data integration and data exploration services designed to facilitate the Phase 1 preprocessing of data. We de-