Oracle Data Miner

Oracle Data Miner allows data analysts to build, evaluate, and score models using an intuitive graphical user interface. Oracle Data Miner guides the data analyst through the data mining process. Oracle Data Miner supports the simultaneous building of models and applications; it can automatically generate the Java and/or SQL code associated with the data mining activities. Oracle Data Miner supports building automated and integrated data mining applications.

Oracle Data Mining 10g supports two programming interfaces, a Java interface and a PL/SQL interface. The Java and PL/SQL interfaces are not compatible; for example, you cannot use the Java interface to apply a model built using the PL/SQL interface. Oracle Data Miner builds models and creates results using the ODM Java interface. Oracle Data Miner cannot be used with mining objects created using the PL/SQL interface. In particular, no models created using the PL/SQL interface will be visible in Oracle Data Miner.

Oracle Data Miner replaces the ODM Components (also called Data Mining for Java (DM4J)). For a comparison of Oracle Data Miner and DM4J, see What’s New in Oracle data Miner?.

The rest of this topic summarizes the features and functionality of the Oracle Data Miner.

Guided Data Mining

A Mining Activity provides an "automated checklist" that ensures that all required data preparation is performed and that data mining activities are performed in the correct order. An activity guide ensures that you perform the data preparation required by the algorithm that you are using. A Mining Activity also sets appropriate defaults for mining.

There are two types of mining activities: build and apply.

Data Exploration and Preparation

Oracle Data Miner includes wizards that aid data exploration and data preparation. You can summarize, recode, discretize (bin), and create new data variables.

Model Build, Test, and Apply

Oracle Data Miner includes the following wizards that guide you through data mining tasks:

You can use Oracle Data Miner to test and apply ODM models that were built using a Build Activity or that were created using either of the ODM 10.2 interfaces.

Data Preparation and Management

The Oracle Data Miner supports viewing and managing the high-level objects defined by metadata in ODM (Oracle Data Mining) schemas. You can view and manage

Oracle Data Miner permits you to perform the following data preparation and data management tasks:

Oracle Data Miner User Interface

The Oracle Data Miner user interface consists of two primary panes: a navigator pane and a viewer pane; there is also an Active Tasks pane. For more information about the user interface, see Oracle User Interface.

Publish to OracleBI Discoverer

Tools | Publish to Discoverer Gateway uses OracleBI Discoverer Gateway to publish data mining result in OracleBI Discoverer. You can publish the following results:

You can also compare classification test metrics.

Documentation

Oracle Data Miner and ODM Code Generator are documented primarily in online help. For details, see Where to Find More Information.