5 Installing Oracle Data Miner

This chapter provides instructions for installing Oracle Data Miner, the graphical user interface to Oracle Data Mining.

See Also:

Oracle Data Miner README

About Oracle Data Miner

Oracle Data Miner is a full-featured graphical tool for developing data mining models for Oracle databases. You can deploy the models from within Oracle Data Miner or in SQL-based or Java applications. Using the Oracle Data Miner Code Generator, you can capture the entire process of model build, test, and apply in a PL/SQL package.

The main features of Oracle Data Miner are summarized in Table 5-1.

Table 5-1 Features of Oracle Data Miner

Feature Description

Activities

In Oracle Data Miner, the creation, testing, and scoring of models is performed in units of work called activities. You can use activities to create, test, and apply models using any of the algorithms supported by Oracle Data Mining.

Defaults

Oracle Data Miner supplies reasonable default settings for model builds. However, a range of parameters are also available for advanced users.

Text Mining

Oracle Data Miner supports text mining.

Predictive Analytics

Oracle Data Miner supports predictive analytics operations: predict, explain, and profile.

Publish as Table

You can use the Publish as Database Table feature of Oracle Data Miner to publish mining results in a table or view for use by query and reporting tools.

Code Generator

You can use the Oracle Data Miner Code Generator to create PL/SQL packages that implement mining activites. The Code Generator is a useful tool for producing code to integrate into data mining applications.

Requirements

Oracle Data Miner requires access to an instance of Oracle Database installed with the Data Mining option. The database can be installed locally on your computer or it can be installed on a remote host.

The connecting user ID must have the database and object permissions described in Chapter 4. If you want to use the sample data provided with the sample programs, the SH schema must be installed in the database and the user ID must be enabled as described in "Create a Data Mining Demo User".

Oracle Data Miner does not require Oracle Client, and a Net Service Name for the database connection is not needed. However, if you wish to import data from text files in Oracle Data Miner, SQL*Loader is required. SQL*Loader is installed with Oracle Client. Follow the instructions in "Install Oracle Client" to obtain SQL*Loader.

Oracle Data Miner requires Java JDK version 1.5. To check the version of Java, you can execute the following at the command prompt:

> java -version

Install Oracle Data Miner

To install Oracle Data Miner:

  1. Go to the Oracle Data Miner page on the Oracle Technology Network:

    http://www.oracle.com/technology/products/bi/odm/odminer.html

    You will find the following available for download:

    • Oracle Data Miner

    • Oracle Data Miner Tutorial

    • Oracle Data Miner PL/SQL Code Generator for JDeveloper

    • Oracle Data Miner PL/SQL Code Generator for SQL Developer

  2. Download Oracle Data Miner and unzip to a new, empty directory. The zip file includes a readme.

  3. Download and unzip the tutorial. The zip file includes sample data, a readme, and the tutorial PDF.

  4. If you wish to use the code generators, follow the installation instructions in the Oracle Data Miner readme.

To start Oracle Data Miner:

  1. Navigate to the bin directory in the Oracle Data Miner installation diretory.

  2. Double click odminerw.exe to launch Oracle Data Miner.

    Note:

    The odminer.exe executable, also in the bin directory, displays a console window that can be used for troublshooting.

  3. To connect to an Oracle database, follow the instructions in "Connect From Oracle Data Miner".