Welcome to the Text Transformation Wizard

Oracle Data Mining lets you build classification, regression, anomaly detection, feature extraction, and k-Means clustering models where the input table has one text column. Before you can use a table containing a text column in a mining operation, you must transform the text column appropriately. The way that the text column is prepared depends on how the data will be used.

The text column in a sequence of data mining operations, such as the tables used to build, test, and apply the same model, must be transformed in the same way.

Oracle Data Miner allows one text column in the input for a mining activity.

This wizard prepares a text column used in a mining activity; the wizard extracts features from the text column and creates a context index.

The output of the wizard is two tables:

You can run this wizard as a transformation from Data | Transform | Text , or it can run as part of a mining activity. If the transform runs as part of an activity, you can run it explicitly from a Text step in the activity.

Since the Oracle Data Mining Java interface requires that text be stored in a table and not in a view, the output of this wizard is a table.

For an overview of Oracle Data Mining support for text mining, see Text Mining.

If you do not want to see this page the next time that you launch the wizard, check the Skip this Page Next Time box.

Click Next to continue.