The data mining component for the Pentaho platform will allow data mining processes - in the form of Weka KnowledgeFlow(s) - to be executed in a action sequence, and results made available to other actions.
Some sample action sequences for data mining with Weka.
An example Weka KnowledgeFlow that trains a JRip rule learner, evaluates its performance on test data, and outputs the learned rule set and the results of evaluation.
Configuring an action definition to execute this KnowledgeFlow. The KnowledgeFlow file to execute is specified as a resource. The KnowledgeFlow component to receive output from is the "TextViewer" and the connection type to listen for is "text".
The output of the data mining process as a result of executing the action sequence.