Mining data for extracting information
The emergence of a vibrant global economy has changed the information demands made by businesses of all sizes and types. However, this globalisation of business has increased the need for more continuous analysis and also for managing data in more sophisticated ways, which expose trends and subtle relationships among data. Data mining (DM) is an analytical process designed to explore data in order to reveal newly discovered patterns and connections. These connections can then be used to validate the findings by applying the detected patterns to new subsets of data. An important function of data mining is the production of a model. A model can be descriptive, giving insight on an underlying process or behaviour. Or a model can be predictive, consisting of an equation or set of rules that makes it possible to predict an unseen or unmeasured value from other, known values. The knowledge gained by DM can be input in a Decision Support (DS) system, which helps in facing difficult decisions. There is a current standard emerging for description of data-mining models called Predictive Mark-up Modelling Language (PMML). PMML describes DM models in the Extensible Mark-up Language (XML), which is the universal format for structured documents and data on the Web. This standard simplifies exchange of results and further co-operation between DM and DS. SolEuNet is a network of expert teams from academia and industry, offering tools and expertise for data mining and decision support. Within the SolEuNet project a vendor independent PMML visualizer was developed, allowing the display and better understanding of PMML models. The software does not require expertise on working with PMML models and eliminates the need to purchase and learn an expensive and complex data mining tool. The software can be adapted to specific needs or be enhanced with additional functionality while in future releases, it will also be possible to edit PMML models.
