The operational data from complex processes contain correlations between influencing variables and target values (= quality). Knowledge of these correlations is an important component for advanced quality monitoring and control. In many cases an extensive stock of process data is already available in corporate databases.

Yet to identify such correlations, it is necessary to be able to process the accumulated data by means of suitable data mining techniques. The identified correlations can then be used to support operating crews in controlling technical processes and in predicting product quality. The use of data mining tools to date, however, has mostly necessitated considerable know-how regarding the integrated methods and processing techniques.

DataWizard opens up the possibility to identify important interdependencies and relationships between quality-determining influencing variables and relevant target values through data-based analysis without familiarisation with the field of mathematico-statistical methods. The system operates according to the wizard principle. It guides the user step by step through interactive dialogue and, as it does so, inquires about the main objectives and boundary conditions for the analysis. The pre-processing of the data then takes place automatically with minimum user interaction. The DataWizard subsequently selects, by itself, the optimum method for analysing the supplied process data according to the input wishes of the user, for which purpose efficient techniques from the fields of statistics and computational intelligence are available. All interim results are offered to the user in text or graphical form. At the end of the analysis the results are edited as a report and can then be printed out or exported.
DataWizard can be used as an independent software tool or as an additional module to the DataTools data mining suite offered by the BFI.