Oracle Data Miner 11.1.0.2

Oracle Data Miner 11.1 is the graphical user interface for Oracle Data Mining 11.1; Oracle Data Miner replaces all previous graphical user interfaces to Oracle Data Mining, including Oracle Data Miner 10.1, Oracle Data Miner 10.2, and Data Mining for Java (DM4J). Models built using Oracle Data Miner 10.1 and DM4J cannot be used with Oracle Data Mining 10.2 and 11.1.

Note: You cannot connect to an Oracle 10g Release 2 database using Oracle Data Miner 11.1; to connect to an Oracle 10g Release 2 database, use Oracle Data Miner 10.2.

This document provides a brief overview of new features of Oracle Data Mining 11g and Oracle Data Miner 11.1, along with installation instructions for Oracle Data Miner.

These release notes describe how to install Oracle Data Miner on Mac OS.

Compatibility with Previous Releases of Oracle Data Miner

Models built using Oracle Data Miner 10.1 or Oracle Data Miner 10.2 cannot be used with Oracle Data Miner 11.1.

Oracle Data Miner requires Oracle 11g Release 1; you cannot connect to any other version of Oracle Database. Any patch version of Oracle 11g Release 1 is compatible with Oracle Data Miner.

You can have both Oracle Data Miner 10.2 and Oracle Data Miner 11.1 installed on the same system; each version of Data Miner connects to a different version of Oracle.

Support and Feedback

If you encounter problems when using Oracle Data Miner, you can report them to Oracle Support (requires current product support contract) at Oracle MetaLink.

You can post general comments and suggestions to the Data Mining Discussion Forum on Oracle Technology Network.

Oracle Data Miner 11.1 Overview

Oracle Data Miner supports the new Oracle Data Mining 11g features, as described in What’s New in Oracle Data Mining?.

Oracle Data Mining includes native import/export facilities for moving data mining objects to other schemas. This is often a required step during deployment. Model import/export is not supported by Oracle Data Miner; it is supported by the Oracle Data Mining Java and PL/SQL programmatic interfaces. For information, see Oracle Data Mining Administrator’s Guide.

The rest of this section briefly describes the following new features of Oracle Data Miner 11.1:

Support for the new Generalized Linear Models algorithm for classification and regression
Support for scoping of nested data and enhanced handling of sparse data
Deprecation of the Adaptive Bayes Network classification algorithm

New Algorithms for Classification and Regression

Oracle Data Mining 11g introduces a new algorithm Generalized Linear Models (GLM). GLM supports two mining functions: classification (logistic regression) and regression (linear regression).

Support for Scoping of Nested Data and Enhanced Handling of Sparse Data

Oracle Data Mining supports nested data types for both categorical and numerical data. Multi-record case data must be transformed to nested columns for mining.

Note: In Oracle Data Mining 11g, Decision Tree and O-Cluster algorithms do not support nested data.

Handling of sparse data and missing values has been standardized across algorithms in Oracle Data Mining 11g. Data is sparse when a high percentage (for example, 90%) of the cells are empty but all the values are assumed to be known. This is the case in market basket data; for example in a supermarket, you have thousands of products, but a customer only purchases a small number at any given time. When some cells are empty, and their values are not known, they are assumed to be missing at random. Oracle Data Mining assumes that missing data in a nested column is a sparse representation, and assumes that missing data in a non-nested column is missing at random.

Adaptive Bayes Network Algorithm Deprecated

In Oracle Data Mining 11g, the Adaptive Bayes Network (ABN) classification algorithm is deprecated. Oracle Data Miner 11.1.0.2 does not support building models using the ABN algorithm.

If you need to create a classification model that provides rules for its predictions, a simple solution would be to use the Decision Tree algorithm.

What’s New in Oracle Data Mining?

Oracle Data Mining 11g Release 1 (11.1), the server that Data Miner connects to, includes the following new algorithms and features:

Automatic and embedded data preparation
Data mining models as schema objects
Scoping of nested data and enhancements of handling of sparse data
Generalized Linear Models for classification and regression
Enhancements to Predictive Analytics
SQL prediction enhancements
Java API upgraded to support the new features of Oracle Data Mining 11g

For information about the new features, see Oracle Data Mining Documentation.

Oracle Data Mining Documentation

Oracle Data Mining 11g Release 1 (11.1) documentation is part of the Oracle Database 11g Release 1 (11.1) Documentation Library. To find Oracle Data Mining documentation, view or download the library; then click the Data Warehousing and Business Intelligence link. Oracle Data Mining documentation is also available at Oracle Data Mining Documentation.

Oracle Data Miner Tutorial

The tutorial for Oracle Data Miner 11.1 is available for download at Oracle Data Miner downloads page.

The tutorial is based on Oracle Data Miner 10.2.0.3; some screens for Oracle Data Miner 11.1.0.2 may be slightly different. The tutorial does not describe text mining.

For information about installing Oracle Data Mining 11.1, see Oracle Data Mining Administrator’s Guide.

In addition to the discussion in the tutorial, the online help for Oracle Data Miner contains an example that illustrates code generation for an apply activity.

The online help for Oracle Data Miner contains a text mining tutorial that illustrates basic text mining. To see the tutorial go to Help | Help Contents | Text Mining Tutorial.

How to Start Oracle Data Miner

Start Oracle Data Miner as follows:

On Microsoft Windows systems, double-click data_miner_dir\bin\odminerw.exe, where data_miner_dir is the folder where Oracle Data Miner is installed.
On UNIX and Linux systems, run the script odminer in the directory data_miner_dir/bin, where data_miner_dir is the directory where Oracle Data Miner is installed.
On Mac OS X, open a terminal and run the script odminer in the directory where Oracle Data Miner is installed. For example, if Data Miner is installed in /Applications/odminer, type
```
cd /Applications/odminer/bin
./odminer & 
```

Note: odminer.exe (without the w in its name) displays a console window that can be used for troubleshooting.

Define a Database Connection

When you start Oracle Data Miner for the first time, you must define a database connection.

Note: The user name and password that you specify when you define the connection must satisfy the requirements of Oracle Data Mining. Oracle Data Mining requires a small number of database permissions, plus SELECT access to the tables containing data for analysis. For details, see the Oracle Data Mining Administrator’s Guide.

The first time that you start Oracle Data Miner, a dialog appears asking for the following information:

Connection Name, the name of the connection
User, the name of the Oracle Data Mining user schema where data mining will take place.
Password, the Oracle Data Mining user password.
Save Password, a check box to indicate if the password is saved. If the password is not saved, it must be entered whenever a connection is made.
Host, the system where Oracle Data Mining is installed.
Port, the port number for the connection, for example, 1521.
SID, the SID for the database where Oracle Data Mining is installed, for example, orcl.
Service Name, the service name for the database where Oracle Data Mining is installed.
Note: You must specify either the SID or the Service Name; you cannot specify both.

Click OK when you finish the definition. You are returned to the Choose Connection dialog. You can now select the connection that you just defined from the drop down box.

You may need to contact your Oracle Data Mining DBA for this information.

You can define additional connections and edit existing ones:

To define a new connection, click New in the Choose Connection dialog.
To edit an existing connection, select the connection in the Connection drop down that you want to edit, and click Edit.

If Data Miner is running, you can manage (create, edit, and delete) database connections on the Connections tab of Tools | Preferences.

Oracle Data Miner Install and Uninstall

Installation instructions depend on the target operating system; you can install Oracle Data Miner on the following operating systems:

Microsoft Windows
UNIX and Linux
Mac OS X

Before you can use Oracle Data Miner, you must connect to an appropriate account in an Oracle 11gRelease 1 (11.1) database. Before you can connect, you must install Oracle Data Mining 11g Release 1 and create at least one user account for data mining. For information about how to do this, see Oracle Data Mining Administrator’s Guide and the installation instructions for the platform that you are using.

Before you install Oracle Data Miner make sure that the requirements described in Oracle Data Miner Requirements are satisfied.

Oracle Data Miner Requirements

The following describe Oracle Data Miner requirements. These requirements must be satisfied before you try to build models.

The Data Mining option to Oracle 11g Release 1 (11.1) Enterprise Edition must be installed. The Data Mining option is included by default in Oracle Database Enterprise Edition.
Any user of the Oracle Database must have a user name and password. Oracle Data Mining requires a small number of database privileges, plus SELECT access to the tables containing data for analysis; for a detailed list of privileges, see Required Privileges.
On all platforms, you must have JDK 1.5. If you don't have the correct version of Java installed, you can obtain it from http://www.javasoft.com. For Mac OS, see Oracle Data Miner Requirements for Mac OS X.
The version of the Oracle Data Mining Java API used by Data Miner is 11.1.
If you wish to use the Oracle Data Mining sample programs, you must install them from the Companion CD, as described in Oracle Data Mining Administrator’s Guide.

Oracle Data Miner and Oracle Data Mining do not have to be installed on the same system. For example, you could install Oracle Data Mining on a system running UNIX and Oracle Data Miner on a PC running Microsoft Windows XP.

There are additional requirements for special purposes:

Oracle Data Miner Requirements for Mac OS X
Text Mining Requirements

Required Privileges

Oracle Data Miner requires following system privileges:

CREATE MINING MODEL
CREATE JOB
CREATE TYPE
CREATE SESSION
CREATE TABLE
CREATE SEQUENCE
CREATE VIEW
CREATE SYNONYM
CREATE PROCEDURE
EXECUTE ON ctxsys.ctx_ddl, required for text mining only. If not available, text mining is disabled.
SELECT privileges on the data being mined

The privileges required by Oracle Data Miner are the same as the privileges required for Oracle Data Mining. See Oracle Data Mining Administrator’s Guide for details about how to create a user with these privileges.

CREATE MINING MODEL is a new requirement for Oracle Data Mining 11.1.

CREATE PROCEDURE is satisfied by CREATE ANY PROCEDURE. EXECUTE ON ctxsys.ctx_ddl is satisfied by EXECUTE ANY PROCEDURE.

Oracle Data Miner searches for privileges by searching for direct grants to the user as well as roles assigned to a user. Oracle Data Miner searches one level deep within a role. For example, suppose that the role ANALYST includes the role ODM_USER; if the privileges are granted in the ODM_USER role (which is "one level down" from the role ANALYST), Oracle Data Miner will find them; if the privileges are defined in a role included in ODM_USER, they will not be found.

Oracle Data Miner Requirements for Mac OS X

Oracle Data Miner has the following requirements for Mac OS X:

Mac OS X 10.5 or higher
Java JRE 1.5 or higher. (Mac OS 10.5 comes with Java 1.5 preinstalled.) You can download Java JRE for Mac from http://www.apple.com.

Oracle Data Mining 11.1 cannot be installed directly on Mac OS X at this time. Therefore, you may have to connect to an Oracle database running on some other platform.

One solution it to use Parallels Desktop for Mac to create a Microsoft Windows virtual machine on your MacIntosh, and install the Microsoft Windows version of Oracle 11.1 EE with the Data Mining option on the Windows virtual machine; you can connect to the Oracle database running in the Microsoft Windows virtual machine. Another solution is to connect to Oracle 11.1 EE with the Data Mining Option on some other machine running Microsoft Windows, UNIX, or Linux.

Text Mining Requirements

The following restrictions apply to text mining using Oracle Data Miner:

You can include one text column in a mining operation.
The text column must reside in a table, not a view. A mining activity will create a table, if necessary.
The text column must have data type CLOB, BLOB, BFILE, LONG, VARCHAR2, XMLType, CHAR, RAW, or LONG RAW.
You must have Oracle Text installed on the Server. Oracle Text is required for indexing text columns. Oracle Text is part of Oracle 11g server; the required INDEXTYPE ctxsys.context is part of the Seed Database. Oracle Text is installed by default when you install the Oracle Database. If you explicitly exclude it, you will not be able to use Oracle Data Miner for text mining.

Install on Microsoft Windows

Data Miner does not require an installer. To install Data Miner, you need an unzip tool. If your system does not include an unzip tool, you can download a free, cross-platform unzip tool, Info-Zip, available at http://www.info-zip.org/.

Note: Do not install this Data Miner release into any existing ORACLE_HOME. You will not be able to uninstall it using Oracle Universal Installer.

Follow these steps to install Oracle Data Miner on Microsoft Windows:

Oracle Data Miner on Microsoft Windows requires Java JDK 1.5. To check the version of Java, use the command java -version in a Command Prompt window.
Create a new directory for the new version of Oracle Data Miner. It is not necessary to uninstall any existing versions of Data Miner.
Download odminer110.zip.
Unzip the entire contents of odminer110.zip to the desired Oracle Data Miner root directory, for example, unzip to C:\odminer.
Run (double click) data_miner_dir\bin\odminerw.exe, where data_miner_dir is the folder where Oracle Data Miner is installed. For example, execute C:\odminer\bin\odminerw.exe
Define a connection as described in Define a Database Connection.

Note: odminer.exe (without the w in its name) displays a console window that can be used for troubleshooting.

Install on UNIX or Linux

Note: Do not install this Data Miner release into any existing ORACLE_HOME. You will not be able to uninstall it using Oracle Universal Installer.

Follow these steps to install Data Miner on UNIX or Linux:

Oracle Data Miner on UNIX or Linux requires Java JDK 1.5. To check the version of Java, use the command java -version.
Create a new directory for the new version of Oracle Data Miner. It is not necessary to uninstall any existing versions of Data Miner.
Download odminer110.zip.
Unzip odminer110.zip to the desired Oracle Data Miner root directory; for example, use the following command to unzip the file to the directory odminer in the current working directory using the unzip command unzip odminer110.zip -d odminer. This command creates the directory odminer (in the current working directory) and extracts the archive into it.
To start Oracle Data Miner, run the script odminer in the directory data_miner_dir/bin, where data_miner_dir is the directory where Oracle Data Miner is installed. If the script is not executable, reset the permissions: chmod +x odminer
Define a connection as described in Define a Database Connection.

Install on Mac OS

Data Miner does not require an installer.

Follow these steps to install data Miner on Mac OS X:

Check that you have the required version of Java.
Create a new directory for the new version of Oracle Data Miner. It is not necessary to uninstall any existing versions of Data Miner.
Download odminer110.zip.
Unzip odminer110.zip to the desired Oracle Data Miner root directory; you can use StuffIt Expander or similar software to unzip the downloaded file.
To start Oracle Data Miner, open a terminal and run the script odminer in the directory where Oracle Data Miner is installed. For example, if Data Miner is installed in /Applications/odminer, type
```
cd /Applications/odminer/bin
./odminer & 
```
Define a connection as described in Define a Database Connection.

Data Miner Uninstall

It is not necessary to uninstall existing versions of Data Miner before you install new versions. However, you should unzip different versions of Data Miner into different directories.

If you wish to uninstall Oracle Data Miner on any platform, delete data_miner_dir, the directory where Oracle Data Miner is installed. Make sure that you delete all of the subdirectories of data_miner_dir.

Oracle Data Miner Code Generator Install and Uninstall

Oracle Data Miner includes a wizard and extensions for Oracle JDeveloper and for Oracle SQL Developer that support using generated code in applications. Follow these steps to generate code:

Verify that the Oracle Data Miner PL/SQL Code Generator Extension Requirements are satisfied.
Install the extension as described in Code Generator Extension Install Using a File or Code Generator Extension Install Using the Update Center.
Start the wizard as described in Using the Code Generator Extension.

Oracle Data Miner PL/SQL Code Generator Extension Requirements

The Code Generator Extension requires Oracle JDeveloper 10.1.3.3 or Oracle SQL Developer 1.5.1 (or a patch release of 1.5.1, such as 1.5.2).

The requirements for installing the Oracle Data Miner PL/SQL Code Generator Extension are the same as those for Oracle Data Miner 11.1.

Certain tasks have additional requirements, as follows:

If you plan to compile and execute the generated code, Oracle Data Mining 11.1 must be available in the database where the code executes.
If you generate code that supports the Oracle Workflow API, Oracle Workflow must be installed and certain grants must be made. Generated code works with any version of Oracle Workflow that runs on Oracle 11.1.
If you plan to use JDeveloper or SQL Developer to debug programs using generated code, you must have the following system privileges:
- DEBUG ANY PROCEDURE
- DEBUG CONNECT SESSION

Code Generator Extension Install Using a File

Follow these steps to install the Oracle Data Miner PL/SQL Code Generator Extension for either Oracle JDeveloper or Oracle SQL Developer:

Download ODMrExtJDev.zip for JDeveloper or ODMrExtSQLDev15.zip for SQL Developer and save it into a temporary location; do not save it in either the jdev_install\jdev\extensions or sqldev_install\sqldeveloper\jdev\extensions directory.
Launch JDeveloper or SQL Developer.
Select Help | Check for Updates
Click Next on the Welcome to Check for Updates Wizard page.
In Step 1 of 3: Source, select Install from Local File and then select the Browse and locate ODMrExtJDev.zip for JDEveloper or ODMrExtSQLDev15.zip. Click Next, and finish the wizard.
The Confirm Restart dialog is displayed; select Yes.
The Migrate User Settings dialog is displayed; select No.
JDeveloper or SQL Developer is restarted with the Data Mining Code Generator Extension enabled.

Code Generator Extension Install Using the Update Center

Note: If the extensions that you wish to install are not in the update center, install them using the directions in Code Generator Extension Install Using a File.

Follow these steps to install the extension using the Update Center:

Launch JDeveloper or SQL Developer.
Select Help | Check for Updates
Click Next on the Welcome to Check for Updates Wizard page.
In Step 1 of 3: Source, select the Official Oracle Extensions and Updates update center, and click Next.
Select Oracle Data Mining PL/SQL Package.
Follow the directions in the wizard to install the extension. The installation will require a restart of JDeveloper or SQL Developer.
The Confirm Restart dialog is displayed; select Yes.
The Migrate User Settings dialog is displayed; select No.
JDeveloper or SQL Developer is restarted with the Data Mining Code Generator Extension enabled.

Using the Code Generator Extension

To verify that the latest version of the Code Generator extension is installed, select Help | About in JDeveloper or SQL Developer and look for the following entry:

Oracle Data Mining PL/SQL Package oracle.dmt.dm4j.extension.codegenerator Version Number Loaded

Follow these steps to generate code for an already-built activity:

In JDeveloper connect to a database, select a schema, and select File | New. Data Mining PL/SQL Package is listed in the New Gallery in the Database Tier under Database Objects.
In SQL Developer, select a connection to a Data Mining Server, connect to it, and then select File | New | Database Tier | Database Objects; Data Mining PL/SQL Package is one of the items.
In either case, click OK to launch the code generator wizard.

For a brief tutorial illustrating how to generate and execute code for an apply activity, see the Code Generation Example in Oracle Data Miner Help Contents. For more examples, see chapter 15 of the Oracle Data Miner Tutorial.

Code Generator Extension Disable

Follow these steps to disable Oracle Data Miner PL/SQL Code Generator Extension from either Oracle SQL Developer or JDeveloper:

Go to the Tools menu and select Preferences.
Select Extensions in the tree selection on the left of the Preferences Dialog.
Unselect (uncheck) Data Mining PL/SQL Package and click OK.
The Confirm Restart dialog will be displayed; select Yes.
To enable again, simply repeat these steps, but select (check) the extension.

Oracle Data Miner Notes

The following notes apply to Oracle Data Miner:

Oracle Data Mining 11g Release 1 supports two interfaces, a Java interface and a PL/SQL interface. The Oracle Data Mining 11g Java and PL/SQL interfaces are compatible; for example, you can use the Java interface to apply a model built using the PL/SQL interface. Oracle Data Miner can be used with mining objects created using either the Oracle Data Mining 11g Java or the Oracle Data Mining 11g PL/SQL interface.
File Import requires SQL*Loader. If you install the Administrator installation type of Oracle Administrative Client, SQL*Loader is installed. SQL*Loader is usually at C:\app\user\product\11.1.0\client_client_num\BIN\sqlldr.exe. You specify the location of SQL*Loader in Tools | Preferences in Data Miner.
The user name that you specify when you connect must be the name of a database user account with the appropriate permissions. See Oracle Data Mining Administrator’s Guide for information about how to create such accounts.
Data Miner ships in English only.
Do not delete any tables or views with names that start with DM4J$. All such tables and views contain metadata used by Oracle Data Miner. If you delete any of these tables or views, Oracle Data Miner will not function and no recovery is possible. In addition, Oracle Data Mining creates tables in the user's schema with names that start with DM$; do not delete any of these files, either.
On high resolution wide monitors, you may find fonts problems. If this is the case, try replacing the JRE packaged with Data Miner with a more recent JRE, as follows:
1. Download the latest version of Java Runtime Environment (JRE) 6 from http://developers.sun.com/downloads/top.jsp. Install the downloaded file.
2. Go to data_miner_dir, the directory where Data Miner is installed.
3. Rename the jre directory to jrebackup.
4. Copy the downloaded JRE to data_miner_dir. If you installed the downloaded JRE to the default location for Microsoft Windows, the JRE directory is C:\Program Files\Java\jre1.6.0_rel, for example, C:\Program Files\Java\jre1.6.0_07, if you downloaded version 7.
5. Rename the copied JRE folder to jre, for example, rename jre1.6.0_07 to jre.

Oracle Data Miner Bugs Fixed in This Release

The following Oracle Data Miner bugs are fixed in this version:

Bug 8507088: TRANSFORM WIZARD: STANDALONE (OUTSIDE OF ACTIVITY)
Bug 8463935: FILTER OUT ODMR$ TABLES GENERATED BY ODMR 11.2
Bug 8364132: FILE IMPORT: MISSING PASSWORD DUE TO PASSWORD NOT SAVED
Bug 7490498: MODEL BUILD WIZARD: UNSELECTED A COLUMN RESULTED IN A QUERY FAILURE
Bug 7486529: CLUSTERING BUILD WIZARD: EXCLUDE ALL BUTTON IN DATA STEP NOT WORKING PROPERLY
Bug 7420614: CODE GEN: NOT DROPPING TEMP TABLES WHEN CODE IS EXECUTED

Oracle Data Miner Bugs

The following are Oracle Data Miner bugs and Oracle Data Mining bugs that affect Oracle Data Miner:

Performance issue with displaying histograms on a large data set using Data Summarization Viewer. This can be avoided by either sampling the data or viewing the data within the Build Mining Activity.
The Mapping Name used for transaction column data mapping cannot be a mixed-case name.
O-Cluster does not support VARCHAR2 case IDs.
O-Cluster and Decision Tree do not support nested columns.
Anomaly Detection model build fails if the nested column name is lowercase.
Naive Bayes model build fails if the data contains a nested column with embedded quote.

Oracle, JD Edwards, PeopleSoft, and Siebel are registered trademarks of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners.