Data Mining INST

Oracle® Data Mining
Administrator's Guide
11g Release 2 (11.2)
E16807-08
July 2013
Oracle Data Mining Administrator's Guide, 11g Release 2 (11.2)
E16807-08
Copyright © 1996, 2013, Oracle and/or its affiliates. All rights reserved.
Primary Author: Kathy L. Taylor
This software and related documentation are provided under a license agreement containing restrictions on
use and disclosure and are protected by intellectual property laws. Except as expressly permitted in your
license agreement or allowed by law, you may not use, copy, reproduce, translate, broadcast, modify, license,
transmit, distribute, exhibit, perform, publish, or display any part, in any form, or by any means. Reverse
engineering, disassembly, or decompilation of this software, unless required by law for interoperability, is
prohibited.
The information contained herein is subject to change without notice and is not warranted to be error-free. If
you find any errors, please report them to us in writing.
If this is software or related documentation that is delivered to the U.S. Government or anyone licensing it
on behalf of the U.S. Government, the following notice is applicable:
U.S. GOVERNMENT END USERS: Oracle programs, including any operating system, integrated software,
any programs installed on the hardware, and/or documentation, delivered to U.S. Government end users
are "commercial computer software" pursuant to the applicable Federal Acquisition Regulation and
agency-specific supplemental regulations. As such, use, duplication, disclosure, modification, and
adaptation of the programs, including any operating system, integrated software, any programs installed on
the hardware, and/or documentation, shall be subject to license terms and license restrictions applicable to
the programs. No other rights are granted to the U.S. Government.
This software or hardware is developed for general use in a variety of information management
applications. It is not developed or intended for use in any inherently dangerous applications, including
applications that may create a risk of personal injury. If you use this software or hardware in dangerous
applications, then you shall be responsible to take all appropriate fail-safe, backup, redundancy, and other
measures to ensure its safe use. Oracle Corporation and its affiliates disclaim any liability for any damages
caused by use of this software or hardware in dangerous applications.
Oracle and Java are registered trademarks of Oracle and/or its affiliates. Other names may be trademarks of
their respective owners.
Intel and Intel Xeon are trademarks or registered trademarks of Intel Corporation. All SPARC trademarks
are used under license and are trademarks or registered trademarks of SPARC International, Inc. AMD,
Opteron, the AMD logo, and the AMD Opteron logo are trademarks or registered trademarks of Advanced
Micro Devices. UNIX is a registered trademark of The Open Group.
This software or hardware and documentation may provide access to or information on content, products,
and services from third parties. Oracle Corporation and its affiliates are not responsible for and expressly
disclaim all warranties of any kind with respect to third-party content, products, and services. Oracle
Corporation and its affiliates will not be responsible for any loss, costs, or damages incurred due to your
access to or use of third-party content, products, or services.
Contents
Preface ................................................................................................................................................................. v
Audience....................................................................................................................................................... v
Documentation Accessibility ..................................................................................................................... v
Related Documentation .............................................................................................................................. vi
Where to Find Database Tuning Documentation .................................................................................. vi
Conventions ................................................................................................................................................. vi
What's New in Oracle Data Mining Administration? ............................................................... ix

Oracle Database 11g Release 2 (11.2.0.3) Oracle Data Mining .............................................................. ix
Oracle Database 11g Release 2 (11.2.0.2) New Features in Oracle Data Mining ................................ ix
Oracle Database 11g Release 1 New Features in Oracle Data Mining ................................................ x
1 Getting Started with Oracle Data Mining

Step One: Install the Database .............................................................................................................. 1-1
Step Two: Verify the Data Mining Installation ................................................................................. 1-2
Step Three: Create a Data Mining User ............................................................................................... 1-2
Step Four: If You Want to Use Oracle Data Miner............................................................................. 1-3
Step Five: If You Want to Use the Spreadsheet Add-In.................................................................... 1-3
Step Six: If You Want to Use the Sample Programs .......................................................................... 1-4
2 Installing Oracle Data Mining

Install Oracle Database ........................................................................................................................... 2-1
Verify the Data Mining Installation..................................................................................................... 2-9
Enable or Disable the Data Mining Option........................................................................................ 2-9
Optionally Install Oracle Database Examples................................................................................. 2-10
Optionally Install Oracle Data Miner ............................................................................................... 2-11
Optionally Install the Spreadsheet Add–In ..................................................................................... 2-11
3 Connecting to an Oracle Database

Connect Application Development Tools........................................................................................... 3-1
Connect SQL*Plus.................................................................................................................................... 3-1
Connect SQL Developer ......................................................................................................................... 3-2
Connect Oracle Data Miner.................................................................................................................... 3-3
Oracle Client and Oracle Net ................................................................................................................. 3-3
iii
Connect the Spreadsheet Add–In....................................................................................................... 3-12
4 Users and Privileges for Data Mining

Create a Database User............................................................................................................................ 4-1
Grant Data Mining Privileges................................................................................................................ 4-2
System Privileges for Data Mining....................................................................................................... 4-3
Object Privileges for Mining Models................................................................................................... 4-4
5 Working With Mining Model Schema Objects

Obtaining Information from the Data Dictionary............................................................................. 5-1
Data Definition Language for Mining Models .................................................................................. 5-3
Adding a Comment to a Mining Model............................................................................................... 5-4
Auditing Mining Models........................................................................................................................ 5-4
Exporting and Importing Mining Models........................................................................................... 5-6
Importing From PMML........................................................................................................................ 5-11
6 Upgrading Oracle Data Mining

Data Mining Upgrade Overview........................................................................................................... 6-1
Upgrade Using Database Upgrade Assistant...................................................................................... 6-2
Upgrade Using Export/Import ............................................................................................................... 6-2
Post Upgrade Steps .................................................................................................................................. 6-3
Downgrade Data Mining........................................................................................................................ 6-4
7 The Data Mining Sample Programs

Installation and Setup ............................................................................................................................. 7-1
The PL/SQL Programs ............................................................................................................................. 7-6
The Java Programs.................................................................................................................................... 7-8
The Sample Data ................................................................................................................................... 7-11
A Verifying Your Data Mining Installation

Connect to the Database......................................................................................................................... A-1
Create a Script or Execute the Commands Individually.................................................................. A-1
Command Output ................................................................................................................................... A-3
Index
iv
Preface
This manual explains how to install Oracle Data Mining and prepare the database for
data mining activities.
The preface contains these topics:
■ Audience
■ Documentation Accessibility
■ Related Documentation
■ Where to Find Database Tuning Documentation
■ Conventions
Audience
This guide is intended primarily for users who want to install, configure, and use
Oracle Data Mining on Microsoft Windows for data mining exercises and
demonstrations.
Note for DBAs: DBAs managing production databases that support

Oracle Data Mining should follow standard administrative practices
as described in the Oracle Database Administrator's Guide.
Building data mining models and batch scoring of mining models
tend to put a DSS-like workload on the system, and single-row scoring
tends to put an OLTP-like workload on the system.
Documentation Accessibility
For information about Oracle's commitment to accessibility, visit the Oracle
Accessibility Program website at
http://www.oracle.com/pls/topic/lookup?ctx=acc&id=docacc.
Access to Oracle Support

Oracle customers have access to electronic support through My Oracle Support. For
information, visit http://www.oracle.com/pls/topic/lookup?ctx=acc&id=info or
visit http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs if you are hearing
impaired.
v
Related Documentation
The documentation set for Oracle Data Mining is part of the Oracle Database 11g
Release 2 (11.2) Online Documentation Library. The Oracle Data Mining
documentation set consists of the following documents:
■ Oracle Data Mining Concepts
■ Oracle Data Mining User's Guide
■ Oracle Data Mining Java API Reference (javadoc)
For detailed information about the Oracle Data Mining PL/SQL interface, see Oracle
Database PL/SQL Packages and Types Reference.
For detailed information about the SQL data mining functions, see Oracle Database SQL
Language Reference.
For an introduction to Oracle application development, see Oracle Database 2 Day
Developer's Guide.
Where to Find Database Tuning Documentation

The Oracle Data Mining documentation set does not address the topic of database
tuning for data mining. The Oracle Database Performance Tuning Guide is the best source
of information for tuning Oracle Database (including when the goal is to invoke
mining build/score operations).
The memory tuning parameters PGA_AGGREGATE_TARGET and SGA_TARGET can easily
have the most impact on data mining. The correct setting of PGA is very important for
model building routines, as well as complex queries and batch scoring. From a Data
Mining perspective, the SGA is generally less of a concern, except that real-time
scoring is enabled by having the model loaded into the shared cursor in the SGA, so
this still should be sized accordingly.
See Also:
Oracle Database Performance Tuning Guide for getting started
Oracle Database Performance Tuning Guide for more detail on memory
configuration
In addition, if parallel execution is desired, then there are INIT.ORA parameters that
control this functionality.
See Also: Chapter "Using Parallel Execution" in Oracle Database

VLDB and Partitioning Guide for information about parallel execution
tuning
Conventions
The following text conventions are used in this document:
Convention Meaning
boldface Boldface type indicates graphical user interface elements
associated with an action, or terms defined in text or the
glossary.
italic Italic type indicates book titles, emphasis, or placeholder
variables for which you supply particular values.
vi
Convention Meaning
monospace Monospace type indicates commands within a paragraph,
URLs, code in examples, text that appears on the screen, or
text that you enter.
ORACLE_HOME ORACLE_HOME represents the directory in which Oracle
Universal Installer installs an Oracle product. ORACLE_HOME
contains subdirectories for Oracle software executables and
network files.
vii
viii
What's New in Oracle Data Mining
Administration?
This section summarizes the new features in Oracle Data Mining.

It contains the following topics:
■ Oracle Database 11g Release 2 (11.2.0.3) Oracle Data Mining
■ Oracle Database 11g Release 2 (11.2.0.2) New Features in Oracle Data Mining
■ Oracle Database 11g Release 1 New Features in Oracle Data Mining
Oracle Database 11g Release 2 (11.2.0.3) Oracle Data Mining

■ The Oracle Data Mining Java API is deprecated in this release.
Note: Oracle recommends that you not use deprecated features in

new applications. Support for deprecated features is for backward
compatibility only
■ Oracle Data Mining supports a new release of Oracle Data Miner. The earlier
release, Oracle Data Miner Classic, is still available for download on OTN, but it is
no longer under active development.
To download Oracle Data Miner 11g Release 2, go to:
http://www.oracle.com/technetwork/database/options/odm/dataminerworkflo
w-168677.html
To download Oracle Data Miner Classic, go to:
http://www.oracle.com/technetwork/database/options/advanced-analytics/o
dm/downloads/index.html
Oracle Database 11g Release 2 (11.2.0.2) New Features in Oracle Data

Mining
In Oracle Data Mining 11g Release 2 (11.2.0.2), you can import externally-created data
mining models when they are presented as valid PMML documents. The model must
be of type RegressionModel, either linear regression or binary logistic regression.
PMML is an XML-based standard for representing data mining models.
The IMPORT_MODEL procedure in the DBMS_DATA_MINING package is overloaded with
new syntax to support PMML import. When invoked with this syntax, the IMPORT_
ix
MODEL procedure will accept a PMML document and translate the information into an
Oracle Data Mining model. This includes creating and populating model tables as well
as SYS model metadata.
External models imported in this way will be automatically enabled for Exadata
scoring offload.
See Also:
Oracle Database PL/SQL Packages and Types Reference for information
about DBMS_DATA_MINING.IMPORT_MODEL
Oracle Data Mining Concepts for information about scoring in Oracle
Exadata
Oracle Database 11g Release 1 New Features in Oracle Data Mining

The following features were introduced in Oracle Data Mining 11g Release 1 (11.1):
■ No DMSYS Schema
■ New Database Privilege Required for Creating Models
■ Enhanced Security
■ Mining Models in the Oracle Data Dictionary
■ Features Not Available in Oracle Data Mining 11g Release 1
■ Features Deprecated in Oracle Data Mining 11g Release 1
No DMSYS Schema
Oracle Data Mining 11g is more tightly integrated with Oracle Database than it was in
previous releases. Data Mining metadata and PL/SQL packages have been migrated
from DMSYS to SYS. The DMSYS schema does not exist in Oracle Database 11g fresh
installations.
New Database Privilege Required for Creating Models

In addition to the privileges needed for data mining in previous releases, the CREATE
MINING MODEL privilege is required for creating models. This privilege should be
added to any accounts being upgraded to 11g.
Additional system and object privileges are available for controlling mining activities
in 11g.
See Also: Chapter 4, "Users and Privileges for Data Mining"
Enhanced Security
Security features of Oracle Data Mining are significantly enhanced in 11g. Improved
security for data mining has several aspects:
■ Previously, Oracle Data Mining used DMSYS as its own repository schema. This
necessitated the granting of advanced database privileges to DMSYS, a non-system
account. In 11g, the Oracle Data Mining metadata repository is in SYS, where it is
accessible only to the system DBA.
■ The SQL auditing system can be used to track operations on mining model objects.
■ New system and object privileges for mining model objects are introduced in 11g.
x
See Also:
"Auditing Mining Models" on page 5-4
Chapter 4, "Users and Privileges for Data Mining"
Mining Models in the Oracle Data Dictionary

New catalog views for Data Mining are introduced in Oracle Database 11g:
■ ALL/DBA/USER_MINING_MODELS provides information about mining models.
■ ALL/DBA/USER_MINING_MODEL_ATTRIBUTES provides information about data
columns used to build a mining model.
■ ALL/DBA/USER_MINING_MODEL_SETTINGS provides information about configuration
settings for mining models.
The ALL/DBA/USER_OBJECTS catalog view now identifies mining models.
See Also: "Obtaining Information from the Data Dictionary" in

Chapter 5
Features Not Available in Oracle Data Mining 11g Release 1

The following features are not supported in Oracle Data Mining 11g:
■ DMSYS schema.
■ Database Configuration Assistant (DBCA) is no longer required for adding Oracle
Data Mining metadata to a database. Oracle Data Mining is part of Oracle
Database metadata in Oracle 11g.
■ Oracle Data Mining Scoring Engine, a separately-installable database option in
Oracle 10g, is not available in Oracle 11g. All functionality of the Data Mining
Scoring Engine option is offered in the Data Mining option.
■ Basic Local Alignment Search Tool (BLAST).
Features Deprecated in Oracle Data Mining 11g Release 1

The following features are deprecated in Oracle Data Mining 11g:
■ Adaptive Bayes Network classification algorithm
■ DM_USER_MODELS view is replaced by data dictionary views
■ Several PL/SQL procedures have been deprecated.
■ GET_DEFAULT_SETTINGS
Replaced with data dictionary views: USER/ALL/DBA_MINING_MODEL_SETTINGS
■ GET_MODEL_SETTINGS
Replaced with data dictionary views: USER/ALL/DBA_MINING_MODEL_SETTINGS
■ GET_MODEL_SIGNATURE
Replaced with data dictionary views: USER/ALL/DBA_MINING_MODEL_
ATTRIBUTES
Note: Oracle recommends that you do not use deprecated features in

new applications. Support for deprecated features is for backward
compatibility only.
xi
See Also: Oracle Database PL/SQL Packages and Types Reference
xii
1
Getting Started with Oracle Data Mining
1
The steps described in this chapter explain how to install Oracle Data Mining locally
on your Windows PC or laptop and start up the client interfaces: Oracle Data Miner
and Oracle Spreadsheet Add-In for Predictive Analytics.
This chapter provides summary information. Details are provided in subsequent
chapters of this manual.
Note: The minimum requirements for a working installation of

Oracle Data Mining are:
■ An installation of Oracle Database Enterprise Edition
■ A database user ID that has data mining privileges (see Chapter 4)
■ Step One: Install the Database

■ Step Two: Verify the Data Mining Installation
■ Step Three: Create a Data Mining User
■ Step Four: If You Want to Use Oracle Data Miner
■ Step Five: If You Want to Use the Spreadsheet Add-In
■ Step Six: If You Want to Use the Sample Programs
Step One: Install the Database

Perform a basic installation of Oracle Database Enterprise Edition and create a starter
database.
1. In Windows Control Panel, Administrative Tools, choose Services. Stop any
Oracle services that may be running on your computer.
2. From the Oracle Database installation directory, run SETUP.EXE to start Oracle
Universal Installer. Follow the instructions as specified on the initial pages, then:
■ On the Select Installation Options page, choose Create and Configure a
Database.
■ On the System Class page, choose Desktop Class.
■ On the Typical Install Configuration page, specify the installation directories,
the database name, and the database system passwords.
Getting Started with Oracle Data Mining 1-1

Step Two: Verify the Data Mining Installation
3. The Installer performs prerequisite checks, displays summary information about

the installation, and copies the Oracle executables and the starter database files to
your computer.
4. The Installer configures Oracle Net to enable client connections.
5. The Installer starts Oracle Database Configuration Assistant to create the starter
database.
6. Database Configuration Assistant displays summary information about the starter
database. If you wish to use Oracle sample data or the Data Mining sample
programs, unlock the SH account.
7. Exit the Installer. The starter database is now running, and Oracle Net is ready to
accept client connections.
8. If necessary, restart the Oracle services that you stopped in step 1.
See: "Install Oracle Database" on page 2-1 for more detailed

instructions
Step Two: Verify the Data Mining Installation

Once you have installed Oracle Database and created a starter database, you may wish
to verify that Oracle Data Mining is functioning properly before proceeding further.
In Appendix A, you will find a series of SQL and PL/SQL commands that perform
basic data mining operations. You can type these commands into SQL*Plus or SQL
Developer to make sure that the database is enabled for data mining.
See: Appendix A, "Verifying Your Data Mining Installation"
Step Three: Create a Data Mining User

Create a database user with data mining privileges.
1. Log in to SQL*Plus or SQL Developer as the system user.
2. Execute the following CREATE USER statement. Specify a user name for dmuser and
password for dmpassword.
CREATE USER dmuser IDENTIFIED BY dmpassword
DEFAULT TABLESPACE users
TEMPORARY TABLESPACE temp
QUOTA UNLIMITED ON users;
Commit;
3. Execute these statements to grant data mining privileges to the user.

GRANT CREATE JOB TO dmuser;
GRANT CREATE MINING MODEL TO dmuser; -- required for creating models
GRANT CREATE PROCEDURE TO dmuser;
GRANT CREATE SEQUENCE TO dmuser;
GRANT CREATE SESSION TO dmuser;
GRANT CREATE SYNONYM TO dmuser;
GRANT CREATE TABLE TO dmuser;
GRANT CREATE TYPE TO dmuser;
GRANT CREATE VIEW TO dmuser;
GRANT EXECUTE ON ctxsys.ctx_ddl TO dmuser; -- required for text mining
GRANT SELECT ON data TO dmuser; -- required for mining data that is

not in your schema
1-2 Oracle Data Mining Administrator's Guide

Step Five: If You Want to Use the Spreadsheet Add-In
4. If the user will need to import or export data mining models, grant this additional
privilege.
GRANT CREATE ANY DIRECTORY TO dmuser;
For import/export operations, the user must create a directory object. The user
may also require additional privileges. See "Exporting and Importing Mining
Models" on page 5-6 for details.
See: Chapter 4, "Users and Privileges for Data Mining" for more
detailed instructions
Step Four: If You Want to Use Oracle Data Miner

Oracle Data Miner, an extension to Oracle SQL Developer 3.0, is a graphical user
interface to Oracle Data Mining. Oracle Data Miner uses a workflow paradigm to
perform data mining tasks.
You can use Oracle Data Miner to explore data, build and evaluate multiple mining
models, and apply the models to new data. By building workflows, you can capture
and document the methodology you use to perform a range of mining tasks. You can
save and share workflows.
To install and launch Oracle Data Miner:
1. Go to the following page on the Oracle Technology Network:
http://www.oracle.com/technetwork/database/options/odm/dataminerworkflo
w-168677.html
2. Follow the installation instructions provided on the page. The basic steps can be
summarized as:
a. Install Oracle Database (also described in Step One: Install the Database earlier
in this chapter).
b. Install Oracle SQL Developer 3.0, which includes Oracle Data Miner 11gR2.
c. Install the Oracle Data Miner repository.
3. For further instructions, use the Online Help provided by SQL Developer.
4. For additional assistance, use the Oracle By Example tutorials. Links to the
tutorials are available with the installation instructions at the URL provided in
Step 1.
Note: Oracle Data Miner Classic, the previous release of Oracle Data
Mining, is still available for download on the Oracle Technology
Network.
http://www.oracle.com/technetwork/database/options/advanced-
analytics/odm/downloads/index.html
Oracle Data Miner Classic is no longer under active development.
Step Five: If You Want to Use the Spreadsheet Add-In

The Oracle Spreadsheet Add-In for Predictive Analytics lets you perform certain
limited data mining operations in an Excel spreadsheet.
Getting Started with Oracle Data Mining 1-3

Step Six: If You Want to Use the Sample Programs
To install and launch the Spreadsheet Add-In:

1. Install Oracle Client and create a Net Service Name as described in "Oracle Client
and Oracle Net" on page 3-3.
2. Download the Spreadsheet Add-In from the Oracle Technology Network.
http://www.oracle.com/technetwork/database/options/odm/odm-pred-analytics-addin
-092973.html
3. Open the zip file and extract the contents to the Microsoft Office Library, typically:
C:\Program Files\Microsoft Office\Office\Library
4. Open Excel and click Tools > Add–Ins. Select Oracle Predictive Analytics from
the Add–Ins dialog box. The OraclePA menu is added to the Excel toolbar.
5. From the Add-In menu in Excel, choose Connect.
6. In the Connect (Oracle Database) dialog:
■ Select the Net Service Name that you created.
■ Provide the user name and password of the data mining user.
Click Connect to launch the Spreadsheet Add-In.
See: "Optionally Install the Spreadsheet Add–In" on page 2-11 for

more detailed instructions
Step Six: If You Want to Use the Sample Programs

Follow these steps to install and configure the sample Data Mining programs:
1. Install Oracle Database Examples as described in "Optionally Install Oracle
Database Examples" on page 2-10.
2. The sample programs are copied to the RDBMS\demo subdirectory of the Oracle
home directory. You can find the PL/SQL programs by searching for dm*.sql.
3. Start SQL*Plus and connect to the database as the system user. Run the
dmshgrants script. Specify the full path to the Oracle home directory and the name
of the data mining user.
@ ORACLE_HOME\RDBMS\demo\dmshgrants dmuser
4. Connect to the database as the Data Mining user. Run the dmsh script.
CONNECT dmuser
@ ORACLE_HOME\RDBMS\demo\dmsh
COMMIT;
5. Once you have completed steps 1 through 4, you can run the PL/SQL programs.
See: Chapter 7, "The Data Mining Sample Programs" for more

detailed instructions

2
2 Installing Oracle Data Mining
This chapter is intended for anyone who wishes to install Oracle Data Mining and its
client interfaces on a personal computer or laptop running Microsoft Windows.
Note: For production installations of Oracle Database with Oracle

Data Mining, refer to the installation instructions for your platform.
See the "Installing and Upgrading" page of the Oracle Database
Online Documentation Library:
http://www.oracle.com/pls/db112/homepage
This chapter contains the following sections:

■ Install Oracle Database
■ Verify the Data Mining Installation
■ Enable or Disable the Data Mining Option
■ Optionally Install Oracle Database Examples
■ Optionally Install Oracle Data Miner
■ Optionally Install the Spreadsheet Add–In
Install Oracle Database

Oracle Data Mining is an option to Oracle Database Enterprise Edition. To perform
data mining activities, you must be able to log on to an Oracle database, and your user
ID must have the database privileges described in Chapter 4.
The instructions in this section explain how to perform a basic installation of Oracle
Database Enterprise Edition on Microsoft Windows. Instructions for creating a starter
database are included.
Installing Oracle Data Mining 2-1

Important: These instructions are not intended as a replacement for

Oracle Database installation documentation. If you have questions, if
you encounter problems during the installation, or if you already have
Oracle components installed on your PC, refer to:
■ Oracle Database Installation Guide for Microsoft Windows
For a high level introduction to Oracle Database installation and
configuration, see:
■ Oracle Database 2 Day DBA
Additional documentation is available on the "Installing and
Upgrading" page of the Oracle Database 11g Online Documentation
Library:
http://www.oracle.com/pls/db112/homepage
1. Before you begin the installation, ensure that your computer meets the system
requirements described in Oracle Database Installation Guide for Microsoft Windows.
Note: If your computer uses Dynamic Host Configuration Protocol

(DHCP), you may need to install a loopback adapter to assign a local
IP address to your computer.
See Oracle Database Installation Guide for Microsoft Windows for
instructions.
2. Log on to your computer as a member of the Administrator's group. If you are

unsure if your user ID is in the Administrator's group, check User Accounts in
Windows Control Panel.
3. Stop any Oracle services that may be running on your computer.
In Windows Control Panel, choose Administrative Tools, then Services. Find the
service names that start with "Oracle". Choose Stop for each one.
Note: To stop an Oracle database, first stop the database service

(OracleServiceservice_name), then stop the database listener
(OracleOraDb11g_homeTNSListener).
4. To start the installation, run SETUP.EXE from the Database installation directory.
Oracle Universal Installer opens and displays the Configure Security Options
page. Type your My Oracle Support email address. Click Next.

5. In the Download Software Updates box, type your My Oracle Support credentials
if you want to automatically download software updates. Click Next.
6. On the Select Installation Option page, choose Create and configure a database.

Note: If you choose Install database software only instead of Create

and configure a database, you may choose which database options
you want to include in the installation. If you do not choose the Data
Mining option during installation, you can enable it later. See "Enable
or Disable the Data Mining Option" on page 2-9 for instructions.
7. On the System Class page, choose Desktop Class then Next.

8. On the Typical Install Configuration page, specify the following information about
the installation:
■ Oracle Base — Specify a base directory for Oracle products on this computer.
If this directory does not exist, the Installer will create it.
Oracle base is the root of the Oracle directory tree. It may contain multiple
Oracle home directories. By convention, Oracle base is referred to as ORACLE_
BASE.
■ Software Location — The Installer provides a Software Location under
ORACLE_BASE. This will be the home directory for this installation of Oracle
Database. If this directory does not exist, the Installer will create it. If the
directory does exist, make sure that it is empty.
■ Database File Location — The Installer provides a Database File Location
under ORACLE_BASE. This directory will hold the data files for the starter
database. If this directory does not exist, the Installer will create it. If the
directory does exist, make sure that it is empty.
■ Database Edition — Choose Enterprise Edition.
■ Global Database Name — Specify a unique name for the starter database.
The Global Database name is the fully–qualified name of the database: for
example, dbname.us.oracle.com. Specify a unique name for dbname that is not
more than eight characters.
■ Administrative Password — Specify a password for the database system
accounts SYS and SYSTEM. The password should have at least eight characters
and include both alphabetic and numeric characters.
You will have the opportunity to change the passwords at a later time.
■ Confirm Password — Re–enter the administrative password.

Note: To obtain information about any of the items on this page,

drag the cursor from right to left over the item. A tooltip that
describes the item will appear.
After you have provided the information, click Next.

9. The Installer analyzes your computer to determine if you have the resources
needed for the installation. If the prerequisite checks are successful, click Next. If
not, click Fix and Check Again.
10. On the Summary page, click Install.

11. The Install Product page displays the progress of the installation.
12. Once the software has been copied to your computer, the Database Configuration
Assistant creates and starts the starter database.

13. The Database Configuration Assistant displays summary information about the
starter database.
At the bottom of the page is a note about passwords for the starter database.
Choose Password Management if you wish to reset the administrative passwords
or unlock any other accounts that are available in the database. By default the SYS
and SYSTEM accounts are unlocked.
Note: If you wish to use Oracle sample data or the Oracle Data
Mining sample programs, unlock the SH account.
See Chapter 7 for information about the sample programs.
Click OK to return to the Database Configuration Assistant page. Click OK to

return to Oracle Universal Installer.

Enable or Disable the Data Mining Option
14. On the Finish page, click Close.
15. Now the database is running and ready to accept connections.
16. If you stopped any Oracle services before running Oracle Universal Installer, you
can restart them at this time.
service names that start with "Oracle". Choose Start for each one.
Note: To restart an Oracle database, first restart the listener

(OracleOraDb11g_homeTNSListener), then restart the database service
(OracleServiceservice_name).
Verify the Data Mining Installation

If you wish to verify that Oracle Data Mining is present and functioning properly in
the database, you can execute a few simple commands at the SQL command line. A set
of commands is provided for this purpose in Appendix A.
You can perform this verification without installing or downloading any additional
software.
See: Appendix A, "Verifying Your Data Mining Installation"
Enable or Disable the Data Mining Option

The Data Mining option is enabled by default during installation of Oracle Database.
However, you can choose to disable Data Mining and other database options when
you install the database software. (See step 6 of Install Oracle Database.)
If you wish to enable or disable a database option for an Oracle home, you can use the
command-line utility chopt.
chopt [ enable | disable ] db_option_name
Set db_option_name. to dm for the Data Mining option.

To enable the Data Mining option:
1. Shut down the database.
srvctl stop database -d db_name
2. Stop the database service, OracleServiceSID, using the Services program in

Control Panel.
3. Execute these commands.
cd ORACLE_HOME/bin
chopt enable dm
4. Start the database service, Oracle ServiceSID, using the Services program in
Control Panel.
5. Start up the database.
srvctl start database -d db_name
To disable the Data Mining option:

1. Shut down the database.

Optionally Install Oracle Database Examples
srvctl stop database -d db_name
2. Stop the database service, OracleServiceSID, using the Services program in

Control Panel.
3. Execute these commands.
cd ORACLE_HOME/bin
chopt disable dm
4. Start the database service, Oracle ServiceSID, using the Services program in
Control Panel.
5. Start up the database.
srvctl start database -d db_name
See Also: Oracle Database Installation Guide for Microsoft Windows for
more information about chopt
Optionally Install Oracle Database Examples

The Data Mining sample programs are installed with Oracle Database Examples. The
installation process copies the Data Mining sample programs, along with examples
and demonstrations of other database features, to the \RDBMS\demo subdirectory of
Oracle home.
Note: You only need to install Oracle Database Examples if you wish
to use the sample programs. If you do not wish to install the sample
programs, you can skip this section.
See Also:
Oracle Database Examples Installation Guide for complete installation
instructions
Chapter 7 for information about the Oracle Data Mining sample
programs
To install Database Examples, perform these steps:

1. Ensure that your computer meets the system requirements described in Oracle
Database Examples Installation Guide.

3. To start the installation, go to the Examples installation directory and run

SETUP.EXE.

Optionally Install the Spreadsheet Add–In
Oracle Universal Installer displays the Specify Install Location page. Specify the
Oracle base and Oracle home provided in the database software installation (step
8).
Do not assume that the directories displayed by the Installer are correct.
Click Next.
Optionally Install Oracle Data Miner

Instructions for downloading and installing Oracle Data Miner are available at
http://www.oracle.com/technetwork/database/options/odm/dataminerworkflow-1
68677.html. Additional details are available in the online help for Oracle Data Miner.
Note: Installation of Oracle Data Miner is not required. Oracle Data

Mining supports a comprehensive SQL-based API and a Spreadsheet
Add-In that supports limited data mining capabilities in Microsoft
Excel.
See Oracle Data Mining User's Guide for an introduction to the SQL and
PL/SQL interfaces to Oracle Data Mining.
See Also: Step Four: If You Want to Use Oracle Data Miner in
Chapter 1

The Oracle Spreadsheet Add–In for Predictive Analytics adds predictive analytics
features to Microsoft Excel. Using simple "one click" data mining for PREDICT, EXPLAIN,
PROFILE, and EXPORT/IMPORT operations, you can mine data stored in Excel or in an
Oracle database.
Note: Installation of the Spreadsheet Add-In is not required. Oracle

Data Mining supports a comprehensive SQL-based API as well as a
full-featured graphical interface, Oracle Data Miner.
See Oracle Data Mining User's Guide for an introduction to the SQL and
PL/SQL interfaces to Oracle Data Mining.
See Also: Oracle Data Mining Concepts for information about Oracle
predictive analytics
Spreadsheet Add–In: Requirements

The Spreadsheet Add-In requires the following:
■ Access to an instance of Oracle Database installed with the Data Mining option.
■ A user ID with the privileges described in Chapter 4.
■ Microsoft Excel 2000 or later.
■ Oracle Objects for OLE and Oracle Net Configuration Assistant.
To obtain these components, Oracle Client must be present on the computer where
Excel and the Spreadsheet Add–In are installed. For Oracle Client installation
instructions, see Chapter 3.

■ An Oracle Net Service Name for the database connection. Instructions for creating
an Oracle Net Service Name are provided in Chapter 3.
Spreadsheet Add–In: Install Steps

To install Oracle Spreadsheet Add–In for Predictive Analytics:
1. Ensure that your system meets the requirements described in "Spreadsheet
Add–In: Requirements" on page 2-11.
2. Download the Spreadsheet Add–In from the Oracle Data Mining Web site at
http://www.oracle.com/technetwork/database/options/odm/odm-pred-analyti
cs-addin-092973.html.
3. Open the zip file and extract the files to the Microsoft Office Library directory,
which has a path such as this one:
C:\Program Files\Microsoft Office\Office\Library
The zip file contains the following files:

■ Predictive_Analytics.xla if you have Microsoft Excel 2000, 2002, or 2003
■ Predictive_Analytics_2007.xla if you have Microsoft Excel 2007
■ PAhelp.chm, the help file for the Spreadsheet Add–In
■ A readme (which you can copy to the directory of your choosing).
4. Open Excel and click Tools > Add–Ins. Select Oracle Predictive Analytics from
the Add–Ins dialog box.
Note: If the Spreadsheet Add–In was previously installed on your

computer, uncheck Oracle Predictive Analytics and click OK to close
the Add–Ins dialog. Then reopen the dialog and check the Oracle
Predictive Analytics box.
5. The OraclePA menu is added to the Excel toolbar.

6. Although the Add–In may use data stored in Excel or in Oracle Database, a
connection to Oracle Database is required. For instructions, see "Connect the
Spreadsheet Add–In" on page 3-12.
Start Data Mining in Excel

Although the Spreadsheet Add–In may use data stored in Excel or in Oracle Database,
a connection to Oracle Database is always required. Once you have created the

connection, you can use the PREDICT, EXPLAIN, PROFILE, and export/import operations
provided by the Add–In.
To create the connection and install the Add–In in Excel, follow the instructions in
"Connect the Spreadsheet Add–In" on page 3-12.
Note: The first time you use the Add–In, you will be prompted to
provide the path to the Java executable. Specify the full path of
jdk\bin\java.exe under Oracle home.
For the installation described in "Install Oracle Database" on page 2-1,
the path would be:
C:\app\myuser\product\11.2.0\dbhome_1\jdk\bin\java.exe


3
3 Connecting to an Oracle Database
This chapter explains how to connect to a database to use Oracle Data Mining. It
contains the following sections:
■ Connect Application Development Tools
■ Connect SQL*Plus
■ Connect SQL Developer
■ Oracle Client and Oracle Net
■ Connect the Spreadsheet Add–In
Connect Application Development Tools

SQL*Plus and SQL Developer and other client tools are installed with Oracle Database.
SQL*Plus is a command line tool for transmitting SQL and PL/SQL to the database.
SQL Developer is a graphical development environment for SQL and PL/SQL.
You can use the Data Mining PL/SQL API in either SQL*Plus or SQL Developer.
Within SQL Developer you can use both the Data Miner graphical tool and the Data
Mining PL/SQL API. For information about the PL/SQL API, see Oracle Data Mining
User's Guide.
If you have installed Oracle Database locally on your PC, you can follow these steps to
start SQL*Plus or SQL Developer:
1. From the Windows Start menu, select Oracle – OraDb11g_homen.
Note: If there are n database installations, they will appear as Oracle

– OraDb11g_home1, Oracle – OraDb11g_home2, ...... Oracle –
OraDb11g_homen in the Start menu.
Each of these refers to a separate ORACLE_HOME. Be sure to select the
appropriate one.
2. Choose Application Development.

3. Choose the name of the application.
Connect SQL*Plus
When SQL*Plus starts up it is already connected to the local database. Enter the name
of a user ID that has data mining privileges, as described in Chapter 4. You will be
prompted to enter the password.
Connecting to an Oracle Database 3-1

Connect SQL Developer
Connect SQL Developer

The first time you run SQL Developer, you may be prompted to provide the path to
the Java executable. If so, specify the full path of jdk\bin\java.exe under the Oracle
home directory for the database. For the installation described in Chapter 2, the path
would be:
C:\app\myuser\product\11.2.0\dbhome_1\jdk\bin\java.exe
To define a connection to the local database:

1. Right click Connections and choose New.
2. The New/Select Database Connection dialog appears.
■ For Connection Name, specify a meaningful name that you will remember.
■ For Username/Password, provide a user name and password for the
connection. The user must have data mining privileges as described in
Chapter 4.
■ For Hostname, specify localhost to indicate that the database is hosted on the
local computer.
■ For Port, specify the port where the database will listen to requests from the
client. The default port is 1521.

Oracle Client and Oracle Net
■ For SID (Service Identifier), specify the simple database name. This is the first
component of the global database name described in the installation
instructions in Chapter 2.
Click Connect to connect to the database using the specified credentials.
Connect Oracle Data Miner

Oracle Data Miner is a choice on the Tools menu of SQL Developer.
Oracle Data Miner requires an administrative (SYS) connection for installation. A

separate connection to the database is required for each user that will use the tool.
Instructions for creating these connections are provided in the Oracle Data Miner
Help. For installation and configuration instructions, go to
http://www.oracle.com/technetwork/database/options/odm/dataminerworkflow-1
68677.html.

To connect to an Oracle Database, a client application must identify the database and
present appropriate credentials. The connection mechanism depends on the type of
client and whether it is running locally or remotely. Some connections require the
installation of Oracle Client software.
There are several methods for specifying login credentials. For example, a client may
present login credentials at the time of the connection, as shown below. The system
will prompt for a password and establish a connection.
SQL>CONNECT username @ "hostname[:port][/databasename]"
You may also identify a connection with a predefined Net Service Name.
Oracle Spreadsheet Add–In for Predictive Analytics, whether running locally or
remotely, requires the installation of Oracle Client and the creation of a Net Service
Name.
See Also: "Create a Net Service Name" on page 3-8

Note: SQL Developer and Oracle Data Miner do not require the
installation of Oracle Client.
For details about Oracle Data Miner installation and configuration, see
Oracle Data Miner 11g Release 2 "Getting Started" on the Oracle
Technology Network:
http://www.oracle.com/technetwork/database/options/odm/datam
inerworkflow-168677.html
Additional details are available in the online help for Oracle Data
Miner.
Install Oracle Client

You must install Oracle Client if:
■ You wish to use Oracle Spreadsheet Add-In for Predictive Analytics
■ You do not have a local database and you wish to use SQL*Plus and other tools for
connecting to a remote Oracle database.
Use the following steps to install Oracle Client on a Windows platform:
Note: The following instructions are not intended as a replacement

for Oracle Client and Net Management documentation. For additional
information, refer to:
■ Oracle Database Client Installation Guide for Microsoft Windows
■ Oracle Database Net Services Administrator's Guide
1. Ensure that your computer meets the system requirements described in Oracle
Database Client Installation Guide for Microsoft Windows.

3. Log on to your computer as a member of the Administrator's group. If you are

unsure if your user ID is in the Administrator's group, check User Accounts in
Windows Control Panel.
4. To start the installation, go to the Client installation directory and run SETUP.EXE
to start Oracle Universal Installer.
5. On the Select Installation Type page, choose Administrator.

Click Next.
6. In the Download Software Updates page, provide your My Oracle Support
credentials if you want to automatically download software updates. Click Next.
7. On the Select Product Languages page, select the language for the Oracle Client
user interface.

Click Next.
8. On the Specify Install Location page, provide the path of the Base directory in
which you installed Oracle Database. (See "Install Oracle Database" step 8.)
The Installer provides a Software Location under Oracle Base. This will be the
home directory of the Oracle client that you are installing (Oracle home). If this
directory does not exist, the Installer will create it. If the directory does exist, make
sure that it is empty.

Click Next.
9. The Installer performs prerequisite checks, then displays the Summary page.
Click Install to complete the set up process and begin the installation.
10. On the Install Product page, the Installer displays a progress bar.
When the installation is successfully completed, click Close.

11. Oracle Net Configuration Assistant starts and displays the Welcome page.
Choose Local Net Service Name Configuration, then click Next.

Oracle Net Configuration Assistant determines that Oracle Client can use the Easy
Connect naming method, as described in "Oracle Client and Oracle Net" on
page 3-3. A Net Service Name is not needed. Simply click Next.
12. Oracle Net Configuration Assistant displays summary information about the
connection configuration.
Click the Password Management button at the bottom of the page to unlock any
schemas that you plan. If you wish, you can also reset the system passwords. If
you will use the Oracle Data Mining sample programs and you did not unlock the
SH schema when you installed the database, you should do so now. (See "Install
Oracle Database".)
Click Exit to exit Oracle Net Configuration Assistant, then click Finish to complete
the installation of Oracle Client.
13. On the End of Installation page, click Exit to exit Oracle Universal Installer.
14. If you stopped any Oracle services before running Oracle Universal Installer, you
can restart them at this time.
service names that start with "Oracle". Choose Start for each one.
Note: To restart an Oracle database, first restart the listener

(OracleOraDb11g_homeTNSListener), then restart the database service
(OracleServiceservice_name).
Create a Net Service Name

The instructions in this section explain how to create a Net Service Name. Follow these
instructions if you wish to use Oracle Spreadsheet Add–In for Predictive Analytics.

Note: The following instructions are not intended as a replacement

for Oracle Client and Net Management documentation. For additional
information, refer to:
■ Oracle Database Client Installation Guide for Microsoft Windows
■ Oracle Database Net Services Administrator's Guide
To create a Net Service Name:

1. From the Windows Start menu, select Oracle – OraClient11g_homen.
Note: If there are n installations of Oracle Client, they will appear as

Oracle – OraClient11g_home1, Oracle – OraClient11g_home2, ....
Oracle – OraClient11g_homen in the Start menu. Be sure to select the
appropriate one.
2. Choose Configuration and Migration Tools.

3. Choose Net Configuration Assistant.
4. Oracle Net Configuration Assistant displays the Welcome page.
Choose Local Net Service Name configuration, then Next.

5. On the next page, choose Add to add a new Net Service Name. Then click Next.
6. Provide the Global Database Name in the Service Name text box. (See "Install
Oracle Database" on page 2-1.)

Choose Next.
7. On the next page specify the communication protocol. The default is TCP.
Choose Next.
8. On the next page, choose Yes, perform a test to test the new connection.
The connection will probably fail until you provide a new user ID and password.

Choose Change Login.

Provide the user ID and password of an account that has data mining privileges
on the host.
9. If the user name and password were the only problem with the connection, Oracle
Net Configuration Assistant makes a successful test connection.
Choose Next.

Connect the Spreadsheet Add–In
10. Oracle Net Configuration Assistant reports that the connection configuration was
successful. When you click Next, you are returned to the Welcome page, where
you have the opportunity to perform additional configuration.
Click Finish to exit Oracle Net Configuration Assistant.

Oracle Spreadsheet Add–In for Predictive Analytics implements predictive analytics
("one–click" data mining) within Microsoft Excel. You can use the Spreadsheet Add–In
to analyze data stored in Excel or in Oracle and to import/export data mining models
to/from Oracle Database. Whether the data is stored in Excel or in the database, the
Spreadsheet Add–In requires a database connection.
The Spreadsheet Add–In requires Oracle Client. It uses a Net Service Name to connect
to the database. Even if the database is local, Oracle Client and a Net Service Name are
required.
To connect to Oracle, follow these steps:
1. Install Oracle Client on your computer as described in "Install Oracle Client" on
page 3-4.
2. Create a Net Service Name as described in "Create a Net Service Name" on
page 3-8.
3. Add the Add–In to Excel as described in "Optionally Install the Spreadsheet
Add–In" on page 2-11.
4. From the Add–In menu, choose Connect.
5. The Connect (Oracle Database) dialog appears.

From the list of predefined Net Service Names, choose the connection you want to
use. Provide a user name and password and click Connect to activate the
Spreadsheet Add–In.
See Also: "Optionally Install the Spreadsheet Add–In" on page 2-11


4
4 Users and Privileges for Data Mining
This chapter explains how to create data mining users and control their mining
activities through the use of database privileges.
See Also: Oracle Database 2 Day + Security Guide for information

about Oracle Database security
This chapter contains the following topics:

■ Create a Database User
■ Grant Data Mining Privileges
■ System Privileges for Data Mining
■ Object Privileges for Mining Models
Create a Database User

To create a database user, you must be logged in as a system administrator. You can
use any tool that transmits SQL to the database, for example Database Control,
SQL*Plus, or SQL Developer.
Example: Create a Database User in SQL*Plus

To create a database user in SQL*Plus:
1. From the Windows Start menu, select the Oracle home of the local database.
Note: If there are n database installations, they will appear as Oracle

– OraDb11g_home1, Oracle – OraDb11g_home2, ...... Oracle –
OraDb11g_homen in the Start menu.
Each of these refers to a separate ORACLE_HOME. Be sure to select the
appropriate one.
2. Choose Application Development.

3. Choose SQL*Plus.
4. Log in with system privileges.
Enter user-name: sys / as sysdba
Enter password: password
Users and Privileges for Data Mining 4-1

Grant Data Mining Privileges
5. To create a user named dmuser, type these commands. Specify a password of your
choosing.
CREATE USER dmuser IDENTIFIED BY password
DEFAULT TABLESPACE USERS
TEMPORARY TABLESPACE TEMP
QUOTA UNLIMITED ON USERS;
Commit;
Specify the USERS and TEMP tablespaces. These tablespaces are included in the
starter database that was created during the installation, and they are typically
included in a database by default. (See "Install Oracle Database" on page 2-1). The
USERS and TEMP tablespaces are generally assigned to users that generate examples
and demonstrations.
Note: Tablespaces for Data Mining users should be assigned

according to standard DBA practices, depending on system load and
system resources.
6. To login as dmuser, type the following.

CONNECT dmuser
See Also:
Oracle Database 2 Day DBA for an introduction to creating database
users
Oracle Database SQL Language Reference for the complete syntax of the
CREATE USER statement
Grant Data Mining Privileges

A set of basic privileges is required for data mining. Some may not be required for all
mining activities, however it is prudent to grant them all as a group. The GRANT
statements in Example 4–1 grant these privileges to dmuser.You could execute these
statements in SQL*Plus after logging in with system privileges.
Example 4–1 Privileges Required for Data Mining

GRANT CREATE JOB TO dmuser;
GRANT CREATE MINING MODEL TO dmuser; -- required for creating models
GRANT CREATE PROCEDURE TO dmuser;
GRANT CREATE SEQUENCE TO dmuser;
GRANT CREATE SESSION TO dmuser;
GRANT CREATE SYNONYM TO dmuser;
GRANT CREATE TABLE TO dmuser;
GRANT CREATE TYPE TO dmuser;
GRANT CREATE VIEW TO dmuser;
GRANT EXECUTE ON ctxsys.ctx_ddl TO dmuser; -- required for text mining
GRANT SELECT ON data TO dmuser; -- required for data
that is not in your schema
Additional access rights are required for exporting and importing mining models, as
described in "Exporting and Importing Mining Models" on page 5-6.

System Privileges for Data Mining
Additional system and object privileges described in "System Privileges for Data
Mining" and "Object Privileges for Mining Models" can be used to enable or restrict
particular mining activities.
Note: To create a user for the Data Mining sample programs, you
need to run two configuration scripts as described in "Installation
and Setup" on page 7-1.
System Privileges for Data Mining

A system privilege confers the right to perform a particular action in the database or
to perform an action on a type of schema objects. For example, the privileges to create
tablespaces and to delete the rows of any table in a database are system privileges.
To grant a system privilege, you must either have been granted the system privilege
with the ADMIN OPTION or have been granted the GRANT ANY PRIVILEGE system
privilege.
The system privileges listed in Table 4–1 control operations on mining models.
Table 4–1 System Privileges for Data Mining

System Privilege Allows you to ....
CREATE MINING MODEL Create mining models in your own schema.
CREATE ANY MINING MODEL Create mining models in any schema.
ALTER ANY MINING MODEL Change the name or cost matrix of any mining model in
any schema.
DROP ANY MINING MODEL Drop any mining model in any schema.
SELECT ANY MINING MODEL Apply (score) a mining model in any schema, also
view model details in any schema.
COMMENT ANY MINING MODEL Add a comment to any mining model in any schema. (See
"Adding a Comment to a Mining Model" on page 5-4.)
AUDIT ANY Generate an audit trail for any mining model (or any
object) in any schema. (See "Auditing Mining Models" on
page 5-4.)
Example: Grant System Privileges for Data Mining

The following statements allow dmuser to score models and view model details in any
schema as long as SELECT access has been granted on the data. However, dmuser can
only create models in the dmuser schema.
GRANT CREATE MINING MODEL TO dmuser;
GRANT SELECT ANY MINING MODEL TO dmuser;
The following statement revokes the privilege of scoring or viewing model details in
other schemas. When this statement is executed, dmuser can only perform data mining
activities in the dmuser schema.
REVOKE SELECT ANY MINING MODEL FROM dmuser;
Users and Privileges for Data Mining 4-3

Object Privileges for Mining Models
Object Privileges for Mining Models

An object privilege confers the right to perform a particular action on a specific schema
object. For example, the privilege to delete rows from the SH.PRODUCTS table is an
example of an object privilege.
A user automatically has all object privileges for schema objects in the user's own
schema. A user can grant object privilege on objects in the user's own schema to other
users or roles.
The object privileges listed in Table 4–2 control operations on specific mining models.
Table 4–2 Object Privileges for Mining Models

Object Privilege Allows you to ....
ALTER MINING MODEL Change the name or cost matrix of the specified mining model
object.
SELECT MINING MODEL Apply (score) and view model details of the specified mining
model object.
Example: Grant Object Privileges on Mining Models

The following statements allow dmuser to apply the model testmodel to the sales
table, specifying different cost matrixes with each apply. The user dmuser can also
rename the model testmodel. The testmodel model and sales table are in the sh
schema, not in the dmuser schema.
GRANT SELECT ON MINING MODEL sh.testmodel TO dmuser;
GRANT ALTER ON MINING MODEL sh.testmodel TO dmuser;
GRANT SELECT ON sh.sales TO dmuser;
This statement prevents dmuser from renaming or changing the cost matrix of
testmodel. However, dmuser can still apply testmodel to the sales table.
REVOKE ALTER ON MINING MODEL sh.testmodel FROM dmuser;
See Also: Chapter 5 for information on mining model schema

objects

5
5 Working With Mining Model Schema Objects
In this chapter, you will learn how to retrieve information about mining models from
the data dictionary and perform various operations on mining models.
See Also: Chapter 4, "Users and Privileges for Data Mining"
This chapter contains the following topics:

■ Obtaining Information from the Data Dictionary
■ Data Definition Language for Mining Models
■ Adding a Comment to a Mining Model
■ Auditing Mining Models
■ Exporting and Importing Mining Models
■ Importing From PMML
Obtaining Information from the Data Dictionary

Mining models are database schema objects. They can be queried in the ALL, DBA, and
USER data dictionary views.
The data dictionary views in Table 5–1 reveal information about mining models
created by Oracle Data Mining.
Table 5–1 Mining Model Data Dictionary Views

ALL_ Views DBA_ Views USER_ Views
ALL_MINING_MODELS DBA_MINING_MODELS USER_MINING_MODELS
ALL_MINING_MODEL_ATTRIBUTES DBA_MINING_MODEL_ATTRIBUTES USER_MINING_MODEL_ATTRIBUTES
ALL_MINING_MODEL_SETTINGS DBA_MINING_MODEL_SETTINGS USER_MINING_MODEL_SETTINGS
Obtaining Information about Mining Models

You can query the ALL_MINING_MODELS data dictionary view to obtain information
about all accessible mining model objects.
SQL> DESCRIBE all_mining_models
Name Null? Type
----------------------------------------- -------- ----------------------------
OWNER NOT NULL VARCHAR2(30)
MODEL_NAME NOT NULL VARCHAR2(30)
MINING_FUNCTION VARCHAR2(30)
Working With Mining Model Schema Objects 5-1

Obtaining Information from the Data Dictionary
ALGORITHM VARCHAR2(30)
CREATION_DATE NOT NULL DATE
BUILD_DURATION NUMBER
MODEL_SIZE NUMBER
COMMENTS VARCHAR2(4000)
The information in the COMMENTS column is generated by a SQL COMMENT statement. The
comment provides descriptive information about the model. If no comment was
generated by SQL COMMENT, the COMMENTS column is empty. See "Adding a Comment to
a Mining Model" on page 5-4.
See Also:
Oracle Data Mining Concepts for information about mining functions
and algorithms
Oracle Data Mining User's Guide for more information about ALL_
MINING_MODELS
Obtaining Information about Mining Model Attributes

You can query the ALL_MINING_MODEL_ATTRIBUTES data dictionary view to obtain
information about all accessible mining model attributes.
SQL> DESCRIBE all_mining_model_attributes
Name Null? Type
----------------------------------------- -------- ----------------------------
ATTRIBUTE_NAME NOT NULL VARCHAR2(30)
ATTRIBUTE_TYPE VARCHAR2(11)
DATA_TYPE VARCHAR2(12)
DATA_LENGTH NUMBER
DATA_PRECISION NUMBER
DATA_SCALE NUMBER
USAGE_TYPE VARCHAR2(8)
TARGET VARCHAR2(3)
Note: The attributes listed in this data dictionary view are the
physical columns in the build data that were used to construct the
model. Some or all of these columns should be present for scoring.
These data attributes are referred to as the model signature.
The term attribute is more accurately used to designate the numericals
and categoricals derived from the data attributes for manipulation by
the algorithm. These model attributes may or may not correspond to
data attributes, depending on transformations and on whether or not
the column is nested. The model attributes can be viewed in the model
details returned by the GET_MODEL_DETAILS functions in the DBMS_
DATA_MINING package.

Data Definition Language for Mining Models
See Also:
Oracle Data Mining User's Guide for more information about attributes
MINING_MODEL_ATTRIBUTES
Oracle Database PL/SQL Packages and Types Reference for details about
the GET_MODEL_DETAILS functions
Obtaining Information about Mining Model Settings

You can query the ALL_MINING_MODEL_SETTINGS data dictionary view to obtain
information about all accessible mining model settings.
SQL> DESCRIBE all_mining_model_settings
Name Null? Type
----------------------------------------- -------- ----------------------------
SETTING_NAME NOT NULL VARCHAR2(30)
SETTING_VALUE VARCHAR2(4000)
SETTING_TYPE VARCHAR2(7)
Model settings are specified in a settings table that is used in the model build. The
settings all have default values. The defaults are used when they are not overridden by
settings specified in the settings table, or when there is no settings table.
See Also:
Oracle Database PL/SQL Packages and Types Reference for more
information about model settings
MINING_MODEL_SETTINGS
Data Definition Language for Mining Models

The DDL operations for mining models are described in Table 5–2. The operations are
procedures in the DBMS_DATA_MINING PL/SQL package.
Table 5–2 DDL for Mining Models

DBMS_DATA_MINING
DDL Procedure Description
Add cost matrix ADD_COST_MATRIX Adds a cost matrix to a classification model
Alter reverse expression ALTER_REVERSE_EXPRESSION Alters the reverse transformation expression associated
with a model
Create model CREATE_MODEL Creates a model
Drop model DROP_MODEL Drops a model
Remove cost matrix REMOVE_COST_MATRIX Removes the cost matrix from a classification model
Rename model RENAME_MODEL Renames a model
See: Oracle Database PL/SQL Packages and Types Reference for details

Adding a Comment to a Mining Model
Adding a Comment to a Mining Model

Comments can be used to associate descriptive information with a database object.
You can associate a comment with a mining model using a SQL COMMENT statement.
COMMENT ON MINING MODEL schema_name.model_name IS string;
Note: To add a comment to a model in another schema, you must

have the COMMENT ANY MODEL system privilege.
See Table 4–1, " System Privileges for Data Mining".
To drop a comment, set it to the empty '' string.

The following statement adds a comment to the model DT_SH_CLAS_SAMPLE in your
own schema.
SQL> COMMENT ON MINING MODEL dt_sh_clas_sample IS
'Decision Tree model predicts promotion response';
You can view the comment by querying the catalog view USER_MINING_MODELS.
SQL> COLUMN comments FORMAT a22
SQL> SELECT model_name, mining_function, algorithm, comments FROM user_mining_models;
MODEL_NAME MINING_FUNCTION ALGORITHM COMMENTS

----------------- ---------------- -------------- -----------------------------------------------
DT_SH_CLAS_SAMPLE CLASSIFICATION DECISION_TREE Decision Tree model predicts promotion response
To drop this comment from the database, issue the following statement:
SQL> COMMENT ON MINING MODEL dt_sh_clas_sample '';
Auditing Mining Models

The SQL auditing system is a DBA tool for tracking operations on database objects in a
production environment. Data mining models can be tracked with the SQL auditing
system.
Note: To audit a mining model in another schema, you must have

the AUDIT ANY system privilege.
See Table 4–1, " System Privileges for Data Mining".
Enabling Auditing in the Database

DBAs use auditing to monitor activity in the database. Audit records can be generated
to monitor various database operations. They include information such as the
operation that was audited, the user performing the operation, and the date and time
of the operation.
Audit records can be stored in either a data dictionary table, called the database audit
trail, or in operating system files, called an operating system audit trail.
The database initialization parameter AUDIT_TRAIL controls auditing capabilities in the
database. To enable auditing, set AUDIT_TRAIL to one of the following:
DB
DB,EXTENDED

Auditing Mining Models
OS
XML
XML,EXTENDED
Set AUDIT_TRAIL to NONE to prevent auditing information from being recorded. By

default, AUDIT_TRAIL is set to DB.
See Also: Oracle Database Security Guide for details about the AUDIT_
TRAIL settings
Opening an Audit Trail on Mining Models

Use the SQL AUDIT statement to open an auditing trail on a data mining model.
AUDIT {operation|ALL} ON MINING MODEL schema_name.model_name;
You can track the following operations on mining models.
Audit Operation Description

AUDIT Generate an audit trail for a mining model
COMMENT Add a comment to a mining model
GRANT Give permission to a user to access the model
RENAME Change the name of the model
SELECT Apply the model or view its signature.
For example, this statement generates an audit trail for all GRANT operations on the
model NB_SH_CLAS_SAMPLE in the DMUSER schema.
SQL> AUDIT GRANT ON MINING MODEL dmuser.nb_sh_clas_sample;
This statement generates an audit trail for all operations on the same model.
SQL> AUDIT GRANT,AUDIT,COMMENT,RENAME,SELECT
ON MINING MODEL dmuser.nb_sh_clas_sample;
You can refine the criteria for auditing with the following additional semantics.
AUDIT {operation|ALL} ON MINING MODEL schema_name.model_name
[BY [SESSION|ACCESS]]
[WHENEVER [NOT] SUCCESSFUL]];
Closing the Audit Trail

Use the NOAUDIT statement to stop one or more auditing operations previously enabled
by the AUDIT statement.
NOAUDIT {operation| ALL} ON MINING MODEL model_name
[WHENEVER [NOT] SUCCESSFUL]];
Viewing the Audit Trail

For each audited operation, Oracle Database produces an audit record containing:
■ The name of the user performing the operation
■ The type of operation
■ The object involved in the operation

Exporting and Importing Mining Models
■ The date and time of the operation

Several data dictionary views present auditing information. Some examples are:
■ DBA_AUDIT_OBJECT displays audit trail records for all objects in the database.
■ USER_AUDIT_OBJECT displays audit trail records for all objects accessible to the
current user
■ DBA_OBJ_AUDIT_OPTS describes auditing options for all objects in the database.
■ USER_OBJ_AUDIT_OPTS describes auditing options for all objects owned by the
current user.
Note: The Oracle Database auditing system is a powerful, highly

configurable tool for tracking operations on schema objects. Refer to
the following manuals for more information:
■ Oracle Database Security Guide for a comprehensive discussion of
database auditing
■ Oracle Database SQL Language Reference for documentation of the
AUDIT and NOAUDIT statements
■ Oracle Database Reference for documentation of the AUDIT_TRAIL
initialization parameter and the data dictionary views for
querying the database audit trail

You can export data mining models to flat files to back up work in progress or to move
models to a different instance of Oracle Database Enterprise Edition (such as from a
development database to a production database). All methods for exporting and
importing models are based in Oracle Data Pump technology.
Using Oracle Data Pump

Oracle Data Pump consists of two command-line clients and two PL/SQL APIs. The
command-line clients, EXPDP and IMPDP, provide an easy-to-use interface to the Data
Pump export and import utilities. EXPDP and IMPDP can be used to export/import
entire schemas or databases.
The Data Pump export utility writes the schema objects, including the tables and
metadata that constitute mining models, to a dump file set. The Data Pump import
utility retrieves the schema objects, including the model tables and metadata, from the
dump file set and restores them in the target database.
EXPDP and IMPDP cannot be used to export/import individual mining models.
See Also: Oracle Database Utilities for information about Oracle

Data Pump and the expdp and impdp utilities
Using EXPORT_MODEL and IMPORT_MODEL

The DBMS_DATA_MINING package includes the EXPORT_MODEL and IMPORT_MODEL
procedures for exporting and importing individual mining models. EXPORT_MODEL and
IMPORT_MODEL use the export and import facilities of Oracle Data Pump.

See Also: Oracle Database PL/SQL Packages and Types Reference

for detailed information about the export and import procedures
in the DBMS_DATA_MINING package
Database Privileges for Export/Import

Your options for exporting and importing mining models depend on your access
rights in the database:
■ A DBA can use EXPDP to export a full database and IMPDP to import a full database.
All mining models in the database are included in the export/import.
■ A DBA or individual user can use EXPDP to export a schema and IMPDP to import a
schema. All mining models in the schema are included in the export/import.
■ A DBA or individual user can use DBMS_DATA_MINING.EXPORT_MODEL to export
individual models and DBMS_DATA_MINING.IMPORT_MODEL to import individual
models. These procedures can export/import a single mining model, all the
mining models in a schema, or mining models that match specific criteria.
By default, EXPORT_MODEL and IMPORT_MODEL support export and import within the
same schema. You may need special privileges in the database to import models
into a different schema. These privileges are granted by the EXP_FULL_DATABASE
and IMP_FULL_DATABASE roles, which are only available to privileged users (such
as SYS or a user with the DBA role). You do not need these roles to export/import
models within your own schema.
To import models, you must have the same database privileges as the user who
created the dump file set. Otherwise, a DBA with full system privileges must
import the models.
Directory Object for EXPORT_MODEL and IMPORT_MODEL

EXPORT_MODEL and IMPORT_MODEL use a directory object to identify the location of the
dump file set. A directory object is a logical name in the database for a physical
directory on the host computer.
To export Data Mining models, you must have write access to the directory object and
to the file system directory that it represents. To import Data Mining models, you must
have read access to the directory object and to the file system directory. Also, the
database itself must have access to file system directory.
You must have the CREATE ANY DIRECTORY privilege to create directory objects.
The following SQL command creates, or re-creates if it already exists, a directory object
named dmtest. The file system directory (in this example,
C:\oracle\product\11.1.0\dmining) must already exist and have shared read/write
access rights granted by the operating system.
CREATE OR REPLACE DIRECTORY dmtest AS 'c:\oracle\product\11.1.0\dmining';
This SQL command gives user dmuser both read and write access to dmtest.
GRANT READ,WRITE ON DIRECTORY dmtest TO dmuser;
Note: For security reasons, use caution in granting permissions.
See Also: Oracle Database SQL Language Reference for information

about the CREATE DIRECTORY statement

Tablespace for IMPORT_MODEL

Mining models are stored in the default tablespace of the mining model owner, or in a
tablespace to which the owner has access. The tablespace must also exist in the target
database, and the target user must have access to it. If the tablespace does not exist in
the target database, you must create it before importing the models.
For example, if the models were created in schema DMUSER and the default tablespace
for DMUSER is USERS, then the USERS tablespace must exist in the target database. You
can create the USERS tablespace and grant access to a target user with appropriate
tablespace quota as follows.
connect / as sysdba;
create tablespace USERS datafile 'data_file_name' size 200M autoextend on;
alter user target_user quota unlimited on USERS;
Tables Created By EXPORT_MODEL and IMPORT_MODEL

The Data Mining export and import utilities create tables in the user's schema that are
for internal use only:
■ DM$P_MODEL_EXPIMP_TEMP. Used for internal purposes during export and import,
and provides a job history.
■ DM$P_MODEL_IMPORT_TEMP. Used only for internal purposes during import.
■ DM$P_MODEL_TABKEY_TEMP. Used only for internal purposes during export and
import.
Do not alter these tables. However, you may drop them when no export or import job
is running. The utilities will re-create them for the next job.
Examples: Exporting and Importing Mining Models

This example creates a dump file containing two models and imports the models from
the dump file.
This example was generated on a Linux system. The directory object mydir identifies
the path /scratch/dumpfiles. (See "Directory Object for EXPORT_MODEL and
IMPORT_MODEL" on page 5-7.)
Exporting Models from the DMUSER Schema

In this example, the dmuser schema contains two mining models.
SQL> CONNECT dmuser
SQL> SELECT model_name FROM user_mining_models;
MODEL_NAME
------------------------------
GLMR_SH_REGR_SAMPLE
SVMC_SH_CLAS_SAMPLE
The following commands create the directory object mydir and export all models from
dmuser to mydir.
SQL> CREATE OR REPLACE DIRECTORY mydir AS '/scratch/dumpfiles';
SQL> EXECUTE dbms_data_mining.export_model('all_dmuser_models.dmp', 'mydir');

An export or import creates a log file in the same directory as the dump file. Error
messages are returned to the current output device (such as the screen), and the log file
may provide additional information.
This sample export created two files in the mydir directory:
■ A dump file named ALL_DMUSER_MODELS01.DMP (note the 2-digit suffix added to
the name)
■ A log file with the name DMUSER_EXP_920.LOG
For detailed information about the default names of files, see Oracle Database PL/SQL
Packages and Types Reference.
You can view the log file using a system command or editor. You must know the path
of the physical directory in order to locate the file.
DMUSER_EXP_920.LOG lists the two exported mining models and supporting objects .
Starting "DMUSER"."DMUSER_exp_17": DM_EXPIMP_JOB_ID=17
Estimate in progress using BLOCKS method...
Processing object type TABLE_EXPORT/TABLE/TABLE_DATA
Total estimation using BLOCKS method: 1.062 MB
Processing object type TABLE_EXPORT/TABLE/TABLE
Processing object type TABLE_EXPORT/TABLE/GRANT/OWNER_GRANT/OBJECT_GRANT
Processing object type TABLE_EXPORT/TABLE/INDEX/INDEX
Processing object type TABLE_EXPORT/TABLE/CONSTRAINT/CONSTRAINT
Processing object type TABLE_EXPORT/TABLE/INDEX/STATISTICS/INDEX_STATISTICS
Processing object type TABLE_EXPORT/TABLE/STATISTICS/TABLE_STATISTICS
Processing object type TABLE_EXPORT/TABLE/POST_INSTANCE/PROCACT_INSTANCE
. . exported "DMUSER"."DM$PIGLMR_SH_REGR_SAMPLE" 7.085 KB 14 rows
. . exported "DMUSER"."DM$PISVMC_SH_CLAS_SAMPLE" 7.507 KB 17 rows
. . exported "DMUSER"."DM$PCGLMR_SH_REGR_SAMPLE" 53.27 KB 2278 rows
. . exported "DMUSER"."DM$PAGLMR_SH_REGR_SAMPLE" 5.796 KB 24 rows
. . exported "DMUSER"."DM$PBGLMR_SH_REGR_SAMPLE" 15.76 KB 67 rows
. . exported "DMUSER"."DM$PDGLMR_SH_REGR_SAMPLE" 8 KB 66 rows
. . exported "DMUSER"."DM$PDSVMC_SH_CLAS_SAMPLE" 9.023 KB 88 rows
. . exported "DMUSER"."DM$PFGLMR_SH_REGR_SAMPLE" 5.656 KB 10 rows
. . exported "DMUSER"."DM$POSVMC_SH_CLAS_SAMPLE" 5.320 KB 8 rows
. . exported "DMUSER"."DM$PXSVMC_SH_CLAS_SAMPLE" 7.265 KB 77 rows
. . exported "DMUSER"."DM$PZSVMC_SH_CLAS_SAMPLE" 6.164 KB 1 rows
. . exported "DMUSER"."DM$P_MODEL_EXPIMP_TEMP" 5.921 KB 2 rows
Master table "DMUSER"."DMUSER_exp_17" successfully loaded/unloaded
******************************************************************************
Dump file set for DMUSER.DMUSER_exp_17 is:
/scratch/dumpfiles/all_dmuser_models01.dmp
Job "DMUSER"."DMUSER_exp_17" successfully completed at 10:30:19
Importing Models Into the Same Schema

The exported models still exist in DMUSER. In this example, we drop the models before
importing from the dump file. An import will not overwrite an existing model with
the same name.
SQL> EXECUTE dbms_data_mining.drop_model('GLMR_SH_REGR_SAMPLE');
SQL> EXECUTE dbms_data_mining.drop_model('SVMC_SH_CLAS_SAMPLE');
The following command restores all models from the dump file to the DMUSER schema.
SQL> EXECUTE dbms_data_mining.import_model('all_dmuser_models01.dmp', 'mydir');

Importing Models Into a Different Schema

A user with the necessary privileges can load the models from a dump file into a
different schema. The target schema must have the same permissions and have access
to the same tablespace as the schema from which the models were exported.
CREATE USER dmuser2 IDENTIFIED BY password
default tablespace dmuser
temporary tablespace temp
quota unlimited on dmuser;
@ $ORACLE_HOME/RDBMS/demo/dmshgrants dmuser2
The import command, also executed as SYS, loads the two models into the DMUSER2
schema.
SQL> EXECUTE dbms_data_mining.import_model('all_dmuser_models01.dmp', 'mydir',
null, null, null, 'todmuser2', 'dmuser:dmuser2');
A parameter specifies TODMUSER2.LOG as the name of the log file; the .LOG extension is
added automatically to the name. The log file shows the names of the imported
models and supporting metadata.
Master table "SYS"."todmusr2" successfully loaded/unloaded
Starting "SYS"."todmusr4": DM_EXPIMP_JOB_ID=21
Processing object type TABLE_EXPORT/TABLE/TABLE
Processing object type TABLE_EXPORT/TABLE/TABLE_DATA
. . imported "DMUSER2"."DM$PIGLMR_SH_REGR_SAMPLE" 7.085 KB 14 rows
. . imported "DMUSER2"."DM$PISVMC_SH_CLAS_SAMPLE" 7.507 KB 17 rows
. . imported "DMUSER2"."DM$PCGLMR_SH_REGR_SAMPLE" 53.27 KB 2278 rows
. . imported "DMUSER2"."DM$PAGLMR_SH_REGR_SAMPLE" 5.796 KB 24 rows
. . imported "DMUSER2"."DM$PBGLMR_SH_REGR_SAMPLE" 15.76 KB 67 rows
. . imported "DMUSER2"."DM$PDGLMR_SH_REGR_SAMPLE" 8 KB 66 rows
. . imported "DMUSER2"."DM$PDSVMC_SH_CLAS_SAMPLE" 9.023 KB 88 rows
. . imported "DMUSER2"."DM$PFGLMR_SH_REGR_SAMPLE" 5.656 KB 10 rows
. . imported "DMUSER2"."DM$POSVMC_SH_CLAS_SAMPLE" 5.320 KB 8 rows
. . imported "DMUSER2"."DM$PXSVMC_SH_CLAS_SAMPLE" 7.265 KB 77 rows
. . imported "DMUSER2"."DM$PZSVMC_SH_CLAS_SAMPLE" 6.164 KB 1 rows
. . imported "DMUSER2"."DM$P_MODEL_EXPIMP_TEMP" 5.921 KB 2 rows
Processing object type TABLE_EXPORT/TABLE/GRANT/OWNER_GRANT/OBJECT_GRANT
Processing object type TABLE_EXPORT/TABLE/INDEX/INDEX
Processing object type TABLE_EXPORT/TABLE/CONSTRAINT/CONSTRAINT
Processing object type TABLE_EXPORT/TABLE/INDEX/STATISTICS/INDEX_STATISTICS
Processing object type TABLE_EXPORT/TABLE/STATISTICS/TABLE_STATISTICS
Processing object type TABLE_EXPORT/TABLE/POST_INSTANCE/PROCACT_INSTANCE
Job "SYS"."todmusr2" successfully completed at 11:15:26
Exporting and Importing Specific Models

You can export/import a single model, a list of models, or a group of models that
share certain characteristics. For example, the following command exports two
models, NMF_SH_SAMPLE and SVMR_SH_REGR_SAMPLE to the directory specified by
directory object EXP_OUT.
SQL>execute dbms_data_mining.export_model ( 'models2_out', 'EXP_OUT',
'name in (''SVMR_SH_REGR_SAMPLE'')');
See: Oracle Database PL/SQL Packages and Types Reference for more
examples

Importing From PMML
Importing From PMML

This functionality is available starting with Oracle Database 11g Release 2 (11.2.0.2)
Data Mining.
You can import a mining model represented in Predictive Model Markup Language
(PMML). The model must be of type RegressionModel, either linear regression or
binary logistic regression.
PMML is an XML-based standard specified by the Data Mining Group
(http://www.dmg.org). Applications that are PMML-compliant can deploy
PMML-compliant models that were created by any vendor. Oracle Data Mining
supports the core features of PMML 3.1 for regression models.
See: Oracle Database PL/SQL Packages and Types Reference for more
information about PMML import

Importing From PMML

6
6 Upgrading Oracle Data Mining
This chapter explains how to upgrade or downgrade Oracle Data Mining 11g Release 2
(11.2). It contains the following sections:
■ Data Mining Upgrade Overview
■ Upgrade Using Database Upgrade Assistant
■ Upgrade Using Export/Import
■ Post Upgrade Steps
■ Downgrade Data Mining
Data Mining Upgrade Overview

Oracle Data Mining upgrade is fully integrated with the Oracle Database upgrade
process. Whether you are upgrading from 11.1 release or from 10g releases, Data
Mining models and metadata are upgraded automatically during the upgrade of
Oracle Database.
To upgrade a database, you can use Database Upgrade Assistant (DBUA) or you can
perform a manual upgrade using export/import utilities.
See: Oracle Database Upgrade Guide for complete database upgrade

instructions
Data Mining Model Upgrade

Upgraded models will continue to work as they did in prior releases. New models that
you create in the upgraded environment can make use of the new mining functionality
introduced in the new release.
Note: Models created by the Oracle Data Mining PL/SQL API can be
upgraded from 10.1 to 11g.
Models created by the Oracle Data Mining Java API cannot be
upgraded from 10.1 to 11g. The 10.1 version of the Java API was no
longer supported in Oracle Data Mining 10.2.
Model upgrade from 9.2 release to 11g releases is not supported.
Upgrading Oracle Data Mining 6-1

Upgrade Using Database Upgrade Assistant
Upgrading from 10g

In Oracle Data Mining 10g, Data Mining metadata and PL/SQL packages are stored in
the DMSYS schema. In Oracle Data Mining 11g, DMSYS no longer exists; Data Mining
metadata objects are stored in SYS.
During the upgrade from 10g to 11g, all Data Mining metadata objects and PL/SQL
packages are migrated from DMSYS to SYS. After the upgrade, when you determine that
there is no need to perform a downgrade, set the initialization parameter COMPATIBLE
to 11.2 and drop the DMSYS schema and its associated objects from the upgraded
database as follows:
SQL> CONNECT / AS sysdba;
SQL> DROP USER dmsys CASCADE;
SQL> DELETE FROM sys.exppkgact$ WHERE schema = 'DMSYS';
SQL> SELECT COUNT(*) FROM dba_synonyms WHERE table_owner = 'DMSYS';
If the result is non-zero rows, create and run a SQL script as follows:
SQL> set head off
SQL> spool directory_path/drop_dmsys_synonyms.sql
SQL> SELECT 'Drop public synonym ' ||'"'||SYNONYM_NAME||'";'
FROM DBA_SYNONYMS WHERE TABLE_OWNER = 'DMSYS';
SQL> spool off
SQL> @directory_path/drop_dmsys_synonyms.sql
Once DMSYS is removed, the SYS.DBA_REGISTRY view will no longer list Oracle Data
Mining as a component.
After upgrading to 11g, you can no longer switch to the Data Mining Scoring Engine
(DMSE). The Scoring Engine does not exist in 11g.
See Also: "Oracle Database 11g Release 1 New Features in Oracle

Data Mining" on page ii-x
Upgrading from 11.1

If you upgrade from 11.1 release to 11.2, check for the existence of DMSYS. If the
database was previously upgraded from 10g, DMSYS may still be present. If DMSYS is still
present in the database, you should remove it.
Upgrade Using Database Upgrade Assistant

Oracle Database Upgrade Assistant provides a graphical user interface that guides you
interactively through the upgrade process. To start the Upgrade Assistant:
1. Go to the Windows Start menu and choose the Oracle home directory.
2. Choose the Configuration and Migration Tools menu.
3. Launch the Upgrade Assistant.
On Linux platforms, run the DBUA utility to upgrade Oracle Database.
Upgrade Using Export/Import

If you wish, you can use a less automated approach to upgrading Oracle Data Mining.
You can export the models created in a previous version of Oracle Database and
import them into a new 11g Release 2 (11.2) database.

Post Upgrade Steps
Export/Import Release 11.1 Data Mining Models

To export models from an 11.1 database to a dump file and import them into an 11.2
database, follow the instructions in "Exporting and Importing Mining Models" on
page 5-6.
Export/Import Release 10g Data Mining Models

To export models from a 10g database to a dump file, follow the instructions in
"Exporting and Importing Mining Models" on page 5-6. If any models were created by
the 10.1 Java API, do not include them in the export; they are not supported in 11g.
Before importing the models from the dump file, run the DMEIDMSYS script to create the
DMSYS schema in the 11g database.
SQL>CONNECT / as sysdba;
SQL>@ORACLE_HOME\RDBMS\admin\dmeidmsys.sql
SQL>EXIT;
Note: The TEMP tablespace must already exist in the 11g database.
The DMEIDMSYS script uses the TEMP and SYSAUX tablespaces to create
the DMSYS schema.
To import the dump file into the database:

%ORACLE_HOME\bin\impdp system\<password>
dumpfile=<dumpfile_name>
directory=<directory_name>
logfile=<logfile_name> .....
SQL>CONNECT / as sysdba;
SQL>EXECUTE dmp_sys.upgrade_models('11.0.0');
SQL>ALTER SYSTEM FLUSH SHARED_POOL;
SQL>EXIT;
If you shutdown the database before operating on the upgraded mining models, this
will also flush the shared pool.
After the import is complete, you should drop the DMSYS schema from the database.
Post Upgrade Steps

After upgrading the database, check the upgrade log file and the SYS.DBA_REGISTRY
view to ensure that the upgrade process completed successfully. Also check the DBA_
MINING_MODELS view in the upgraded database. The newly upgraded mining models
should be listed in this view.
After you have verified the upgrade and confirmed that there will be no need to
downgrade, you should set the initialization parameter COMPATIBLE to 11.2.
Important: In Oracle 11g, the CREATE MINING MODEL privilege must

be granted to Data Mining user accounts. This privilege is required for
creating Data Mining models in 11g. Refer to Chapter 4, "Users and
Privileges for Data Mining" for more information.
Upgrading Oracle Data Mining 6-3

Downgrade Data Mining
Downgrade Data Mining

Before downgrading the database back to the previous version, ensure that no 11g
Release 2 (11.2) mining models were created in the upgraded database. Issue the
following SQL statement in SYS to verify:
SQL>SELECT o.name FROM sys.model$ m, sys.obj$ o
WHERE m.obj#=o.obj# AND m.version=2;
If there are any 11g Release 2 (11.2) mining models in the database, you must manually
delete them using the DBMS_DATA_MINING.DROP_MODEL routine before downgrading the
database. If you do not do this, the database downgrade process will be aborted. See
Oracle Database PL/SQL Packages and Types Reference for the calling syntax of DROP_
MODEL.

7
7 The Data Mining Sample Programs
You can learn a great deal about the Oracle Data Mining APIs from the Data Mining
sample programs. The programs illustrate typical approaches to data preparation,
algorithm selection, algorithm tuning, testing, and scoring. Each program creates a
mining model in the database. All the programs include extensive inline comments to
help you understand the code.
See Also: Oracle Data Mining User's Guide for information about the
Data Mining APIs
Note: The Oracle Data Mining Java API is deprecated in this release.
The Java sample programs are still shipped, but Oracle recommends
that you not use the Oracle Data Mining Java API in new applications.
Support for deprecated features is for backward compatibility only
This chapter includes the following sections:

■ Installation and Setup
■ The PL/SQL Programs
■ The Java Programs
■ The Sample Data
Installation and Setup

The Data Mining sample programs are installed with Oracle Database Examples. They
are also available for download from the Oracle Technology Network:
The programs require access to a database that includes the sample schemas. Before
you can run the programs, you must run two configuration scripts to configure the
data and assign the required privileges to your user ID.
Install the Sample Programs

Follow these steps to install the sample programs:
1. Install Oracle Database with the sample schemas, or obtain access to a database
that includes the sample schemas.
■ If you followed the instructions in "Install Oracle Database" on page 2-1, the
sample schemas are installed automatically in the starter database. Be sure to
unlock the SH schema.
The Data Mining Sample Programs 7-1

■ If the database does not include the sample schemas, you can install them
manually or by using Oracle Database Configuration Assistant. See Oracle
Database Sample Schemas for instructions.
2. Determine whether or not Database Examples was installed with Oracle Database.
Database Examples provides a set of sample programs that illustrate numerous
features of Oracle Database, including Oracle Data Mining. The programs are
loaded into the RDBMS/demo subdirectory of Oracle home.
If Database Examples was not installed, you can perform the installation by
following the instructions in "Optionally Install Oracle Database Examples" on
page 2-10. Alternatively, you can download the Data Mining sample programs
from the Oracle Technology Network.
http://www.oracle.com/technetwork/database/options/odm/index.html
Run the Configuration Scripts

Follow these steps to configure the sample data and grant the necessary privileges to
your data mining user ID.
1. Log in to SQL*Plus with system privileges.
Enter user-name: sys / as sysdba
2. If you do not have a user ID for your data mining activities, you can create one by
following the instructions in "Example: Create a Database User in SQL*Plus" on
page 4-1.
3. Run dmshgrants.sql to grant data mining privileges and SH access to your user
ID. Several tables in SH are used by the Data Mining sample programs. Specify the
data mining user name as the parameter. Specify the full path of the Oracle home
directory.
@ ORACLE_HOME\RDBMS\demo\dmshgrants dmuser
4. Now connect to the database as the Data Mining user.

CONNECT dmuser
5. Run dmsh.sql to populate the schema of the Data Mining user with tables, views,
and other objects needed by the sample programs. Specify the full path of the
Oracle home directory.
@ ORACLE_HOME\RDBMS\demo\dmsh
COMMIT;
Locate the Sample Programs

This section explains how to locate the sample programs if they were installed with
Database Examples.
To locate the PL/SQL programs, navigate to the parent directory and search for the
files that start with dm and end with .sql.
For example, if Database Examples was installed in Oracle home
C:\app\demotest\product\11.2.0\db_1\, then navigate to
C:\app\demotest\product\11.2.0\db_1\RDBMS\demo\ and use Windows Search to
find the files named dm*.sql. Windows Search returns the list of Data Mining PL/SQL
programs, as shown in Figure 7–1.

Figure 7–1 The Data Mining Sample PL/SQL Programs
Note: The files listed in Figure 7–1 include all the Data Mining
PL/SQL programs. However, one of the files, dmhpdemo.sql, is not a
Data Mining program.
Use Windows Search to find the files named dm*.java in the same directory. Windows
Search returns the Data Mining Java programs, as shown in Figure 7–2.

Figure 7–2 The Data Mining Sample Java Programs
Run the Sample Programs

You can run the sample programs as many times as you wish. The programs clean up
the results of the previous run before executing the current run.
While the program is running, it displays the program code and the program output.
Run the PL/SQL Sample Programs

To run the PL/SQL programs:
1. Start SQL*Plus and log in as the Data Mining user.
Enter user-name: dmuser
2. Run the program by specifying an at sign (@) followed by the fully-qualified path
of the program. In the following example, replace ORACLE_HOME with the path of
the Oracle home directory.
SQL>@ ORACLE_HOME\RDBMS\demo\dmnbdemo
This example executes the program dmnbdemo.sql, which creates a Naive Bayes
model.
Prepare to Run the Java Programs

Before you can run the Java programs, you must set up your Java environment and
compile the programs.
1. Check that the version of Java you are using is 1.5 or higher. You can execute the
following in a command window to check the version of Java.

>java -version
2. Add ORACLE_HOME\jdk\bin\ to your PATH variable before the paths of any other
Java versions. ORACLE_HOME is the full path to the Oracle home directory.
3. Add the following Data Mining JAR files to your Windows CLASSPATH:
ORACLE_HOME\RDBMS\jlib\jdm.jar
ORACLE_HOME\RDBMS\jlib\ojdm_api.jar
ORACLE_HOME\RDBMS\jlib\xdb.jar
ORACLE_HOME\jdbc\lib\ojdbc5.jar
ORACLE_HOME\oc4j\j2ee\home\lib\connector.jar
ORACLE_HOME\jlib\orai18n.jar
ORACLE_HOME\jlib\orai18n-mapping.jar
ORACLE_HOME\lib\xmlparserv2.jar
4. Compile the programs listed in Figure 7–2. To use the JAVAC executable, open a
command window and go to \RDBMS\demo in Oracle home.
>javac program_name.java
For example:
>javac dmnbdemo.java
If JAVAC is not found, then check the value of the PATH variable.
Run the Java Programs

You can run the Java programs from the operating system prompt with a command
like this:
>java program_name host_name:port_number:database_identifier user password
List the Sample Models

The mining models created by the sample programs can be viewed with a query like
the one shown in Example 7–1.
Example 7–1 Sample Data Mining Models

SQL> SELECT model_name, mining_function, algorithm FROM user_mining_models
ORDER BY model_name;
MODEL_NAME MINING_FUNCTION ALGORITHM

------------------------------ ------------------------------ ------------------------------
ABNMODEL_JDM CLASSIFICATION ADAPTIVE_BAYES_NETWORK
ABN_SH_CLAS_SAMPLE CLASSIFICATION ADAPTIVE_BAYES_NETWORK
AIMODEL_JDM ATTRIBUTE_IMPORTANCE MINIMUM_DESCRIPTION_LENGTH
AI_SH_SAMPLE ATTRIBUTE_IMPORTANCE MINIMUM_DESCRIPTION_LENGTH
APMODEL_JDM CLASSIFICATION NAIVE_BAYES
ARMODEL_JDM ASSOCIATION_RULES APRIORI_ASSOCIATION_RULES
AR_SH_SAMPLE ASSOCIATION_RULES APRIORI_ASSOCIATION_RULES
AR_SH_SAMPLE_STR_XNAL ASSOCIATION_RULES APRIORI_ASSOCIATION_RULES
AR_SH_SAMPLE_XNAL_SVAL ASSOCIATION_RULES APRIORI_ASSOCIATION_RULES
DT_SH_CLAS_SAMPLE CLASSIFICATION DECISION_TREE
GLMCMODEL_JDM CLASSIFICATION GENERALIZED_LINEAR_MODEL
GLMC_SH_CLAS_SAMPLE CLASSIFICATION GENERALIZED_LINEAR_MODEL
GLMRMODEL_JDM REGRESSION GENERALIZED_LINEAR_MODEL
GLMR_SH_REGR_SAMPLE REGRESSION GENERALIZED_LINEAR_MODEL
KMMODEL_JDM CLUSTERING KMEANS

The PL/SQL Programs
KM_SH_CLUS_SAMPLE CLUSTERING KMEANS

NBEXPIMPMODEL_JDM CLASSIFICATION NAIVE_BAYES
NBMODEL_JDM CLASSIFICATION NAIVE_BAYES
NB_SH_CLAS_SAMPLE CLASSIFICATION NAIVE_BAYES
NMFMODEL_JDM FEATURE_EXTRACTION NONNEGATIVE_MATRIX_FACTOR
NMF_SH_SAMPLE FEATURE_EXTRACTION NONNEGATIVE_MATRIX_FACTOR
OCMODEL_JDM CLUSTERING O_CLUSTER
OC_SH_CLUS_SAMPLE CLUSTERING O_CLUSTER
SVMCMODEL_JDM CLASSIFICATION SUPPORT_VECTOR_MACHINES
SVMC_SH_CLAS_SAMPLE CLASSIFICATION SUPPORT_VECTOR_MACHINES
SVMOMODEL_JDM CLASSIFICATION SUPPORT_VECTOR_MACHINES
SVMO_SH_CLAS_SAMPLE CLASSIFICATION SUPPORT_VECTOR_MACHINES
SVMRMODEL_JDM REGRESSION SUPPORT_VECTOR_MACHINES
SVMR_SH_REGR_SAMPLE REGRESSION SUPPORT_VECTOR_MACHINES
TREEMODEL_JDM CLASSIFICATION DECISION_TREE
TXTNMFMODEL_JDM FEATURE_EXTRACTION NONNEGATIVE_MATRIX_FACTOR
TXTSVMMODEL_JDM CLASSIFICATION SUPPORT_VECTOR_MACHINES
T_NMF_SAMPLE FEATURE_EXTRACTION NONNEGATIVE_MATRIX_FACTOR
T_SVM_CLAS_SAMPLE CLASSIFICATION SUPPORT_VECTOR_MACHINES
The model names distinguish the models created by the Java programs from those
created by the PL/SQL programs. The models created by the Java programs have "_
JDM" appended to the name.
The PL/SQL Programs

The PL/SQL sample programs illustrate the use of the DBMS_DATA_MINING package for
creating models and the DBMS_DATA_MINING_TRANSFORM package for performing
transformations on the mining data.
See Also:
■ Oracle Database PL/SQL Packages and Types Reference for syntax of
the PL/SQL API
■ Oracle Data Mining User's Guide for information on the use of the
APIs
■ Oracle Database SQL Language Reference for syntax of the SQL
functions for model scoring
PL/SQL Programs: Algorithms

The PL/SQL programs are presented by algorithm in Table 7–1.
Table 7–1 Algorithms in PL/SQL Sample Programs

Program File Algorithm Mining Function or Task
dmaidemo.sql Minimum Descriptor Length Attribute Importance
dmardemo.sql Apriori Association
dmdtdemo.sql Decision Tree Classification
dmdtxvlddemo.sql Decision Tree (cross validation) Classification
dmglcdem.sql Binary Logistic Regression (GLM) Classification
dmglrdem.sql Multivariate Linear Regression (GLM) Regression
dmkmdemo.sql k-Means Clustering

The PL/SQL Programs
Table 7–1 (Cont.) Algorithms in PL/SQL Sample Programs

dmnbdemo.sql Naive Bayes Classification
dmnmdemo.sql Non-Negative Matrix Factorization Feature Extraction
dmocdemo.sql O-Cluster Clustering
dmsvcdem.sql Support Vector Machine Classification
dmsvodem.sql Support Vector Machine Anomaly Detection
dmsvrdem.sql Support Vector Machine Regression
dmtxtfe.sql Term extraction using Oracle Text Text transformation for
mining
dmtxtnmf.sql Non-Negative Matrix Factorization Text mining using NMF
dmtxtsvm.sql Support Vector Machine Text mining using SVM
PL/SQL Programs: Mining Functions

The PL/SQL sample programs are presented by mining function in Table 7–2. For
detailed descriptions of the sample programs, see the comments in the source code.
Table 7–2 Mining Functions of PL/SQL Sample Programs

Mining Function Description
Classification The classification programs demonstrate various preprocessing
techniques and perform the following steps:
■ Build a classification model using training data
■ Display model details and settings
■ Test the model by applying the model on the test data
■ Compute test metrics, such as confusion matrix, lift, and ROC
■ Apply the model on the scoring data
■ Present apply results
■ Present ranked apply results, influenced by a cost matrix
dmnbdemo.sql illustrates Naive Bayes.
dmdtdemo.sql illustrates Decision Tree.
dmsvcdem.sql illustrates SVM classification.
dmglcdem.sql illustrates GLM classification (binary logistic regression)
The dmdtxvlddemo.sql program demonstrates cross-validation
techniques for decision tree based-classification. With minor
modifications, this program can be used to perform cross validation
using other models/algorithms.
Regression dmsvrdem.sql uses different test metrics, but otherwise performs most
of the same steps used in the classification programs. Selected
attributes of the input data are preprocessed (normalized).
NOTE: dmsvrdem.sql illustrates the use of Automatic Data Preparation.
dmglrdem.sql illustrates GLM regression (multivariate linear
regression)
Anomaly Detection dmsvodem.sql illustrates one-class SVM
Association dmardemo.sql builds an association model and presents frequent
itemsets and association rules as output.

The Java Programs
Table 7–2 (Cont.) Mining Functions of PL/SQL Sample Programs

Mining Function Description
Clustering dmkmdemo.sql (k-Means) and dmocdemo.sql (0-Cluster) build
clustering models and present cluster details, such as rules, centroid,
and histogram for each cluster as output. The models are scored, and
the probabilities associated with each cluster are returned as output.
Selected attributes of the input data are preprocessed.
NOTE: dmkmdemo.sql illustrates the use of Automatic Data Preparation.
Feature extraction dmnmdemo.sql builds a feature extraction model and presents model
details as the output. The model is scored, and each feature ID is
associated with a probability. Selected attributes of the input data are
preprocessed (normalized).
Attribute importance dmaidemo.sql builds an attribute importance model and presents a list
of important attributes as the output of model details. Selected
attributes of the input data are preprocessed (binned).
PL/SQL Text Mining Programs

Three sample programs illustrate the process of text mining using PL/SQL. One
program illustrates the preprocessing that is required to transform the text for mining.
The other two programs build models that use the transformed text.
See Also:
■ Oracle Data Mining Concepts for information about text mining
with Oracle Data Mining
■ Oracle Data Mining User's Guide for details about text
transformation in PL/SQL
The PL/SQL sample text mining programs are:

■ dmtxtfe.sql — Illustrates the process of term extraction that prepares text for
mining
■ dmtxtnmf.sql — Creates a text mining model using the Non-Negative Matrix
Factorization algorithm
■ dmtxtsvm.sql — Creates a text mining model using SVM classification
The Java Programs

The Java demos illustrate the features of the Oracle Data Mining Java API, which
implements Oracle-specific extensions to the Java Data Mining (JDM) 1.0.1.1 standard.
The sample Java programs demonstrate all the Data Mining algorithms as well as data
transformation techniques, predictive analytics, export/import, and text mining.
Note: The Oracle Data Mining Java API is deprecated in this release.
Oracle recommends that you not use deprecated features in new
applications. Support for deprecated features is for backward
compatibility only

The Java Programs
See Also: ■
■ Oracle Data Mining User's Guide for information on the Oracle

Data Mining Java API
■ Oracle Data Mining Java API Reference (javadoc)
Java Programs: Algorithms

The Java programs are presented by algorithm in Table 7–3.
Table 7–3 Algorithms in Java Sample Programs

dmaidemo.java Minimum Description Length Attribute importance
dmapplydemo.java Naive Bayes Scoring methods
dmardemo.java Apriori Association
dmexpimpdemo.java export/import Model Export/Import
dmglcdemo.java Binary Logistic Regression (GLM) Classification
dmglrdemo.java Multivariate Linear Regression (GLM) Regression
dmkmdemo.java k-Means Clustering
dmnbdemo.java Naive Bayes Classification
dmnmdemo.java Non-Negative Matrix Factorization Feature extraction
dmocdemo.java O-Cluster Clustering
dmpademo.java Automated predict and explain Predictive Analytics
dmsvcdemo.java Support Vector Machine Classification
dmsvodemo.java Support Vector Machine (one class) Classification
dmsvrdemo.java Support Vector Machine Regression
dmtreedemo.java Decision Tree Classification
dmtxtnmfdemo.java Non-Negative Matrix Factorization Text mining with NMF
dmtxtsvmdemo.java Support Vector Machine Text mining with SVM
classification
dmxfdemo.java Binning, clipping, and normalization Data Transformations
Java Programs: Mining Functions

The Java sample programs are presented by mining function in Table 7–4. For detailed
descriptions of the sample programs, see the comments in the source code.

The Java Programs
Table 7–4 Mining Functions of the Java Sample Programs

Mining Function or Task Description
Classification The classification programs demonstrate various preprocessing
techniques and perform the following steps:
■ Build a classification model using training data
■ Display model details and settings
■ Test the model by applying the model on the test data
■ Compute test metrics, such as confusion matrix, lift, and
ROC
■ Apply the model on the scoring data
■ Present apply results
■ Present ranked apply results, influenced by a cost matrix
The dmapplydemo.java program demonstrates several ways of
applying a Naive Bayes model.
dmglcdemo.java illustrates GLM classification (binary logistic
regression)
Regression dmsvrdemo.java uses different test metrics, but otherwise
performs most of the same steps used in the classification
programs. Selected attributes of the input data are preprocessed
(normalized).
dmglrdemo.java illustrates GLM regression (multivariate linear
regression)
Association dmardemo.java builds an association model and presents
frequent itemsets and association rules as output. Selected
attributes of the input data are preprocessed (binned).
Clustering dmkmdemo.java (k-Means) and dmocdemo.java (0-Cluster) build
clustering models and present cluster details, such as rules,
centroid, and histogram for each cluster as output. The models
are scored, and the probabilities associated with each cluster are
returned as output. Selected attributes of the input data are
preprocessed (normalized).
Feature extraction dmnmdemo.java builds a feature extraction model and presents
model details as the output. The model is scored, and each
feature ID is associated with a probability. Selected attributes of
the input data are preprocessed (normalized).
Attribute importance dmaidemo.java builds an attribute importance model and
presents a list of important attributes as the output of model
details. Selected attributes of the input data are preprocessed
(binned).
Data transformations dmxfdemo.java demonstrates binning, clipping, and
normalization transformations.
Predictive Analytics dmpademo.java demonstrates PREDICT, EXPLAIN, and PROFILE
functions.
Model import/export dmexpimpdemo.java builds a Naive Bayes model, exports it to a
dump file, then imports it from the dump file.
Java Text Mining Programs

Two Java programs illustrate the process of text mining. One builds a feature
extraction model, the other builds a classification model.

The Sample Data
See Also:
■ Oracle Data Mining Concepts for information about text mining
with Oracle Data Mining
■ Oracle Data Mining User's Guide for details about text
transformation in Java
The Java text mining programs both use the dmtxtnmfdemo.java interface to
transform the text for mining. The programs are as follows:
■ dmtxtnmf.sql — Creates a text mining model using the Non-Negative Matrix
Factorization algorithm
■ dmtxtsvmdemo.java — Creates a text mining model using SVM classification
The Sample Data

The dmsh.sql script creates views, tables, and indexes in the user's schema. The views
define columns of customer data from tables in the SH schema. This data is used by the
Data Mining sample programs. The tables reference the same columns in SH, but they
include an extra COMMENTS column for text mining. The indexes are used to extract
terms from the text in the COMMENTS column and build a nested table column.
Customer Data for Data Mining

Views in the data mining user's schema define columns of data from the CUSTOMERS,
SALES, PRODUCTS, COUNTRIES, and SUPPLEMENTARY_DEMOGRAPHICS tables in the SH
schema. You can list these views with the following SQL statements.
SQL>CONNECT dmuser
SQL>SELECT view_name FROM user_views;
The views are listed in Table 7–5.
Table 7–5 Views Used by the Data Mining Sample Programs

View Name Description
MINING_DATA_APPLY_STR_V Scoring data for o-cluster
MINING_DATA_BUILD_STR_V Training data for o-cluster
MINING_DATA_APPLY_V Scoring data for data mining (not text mining)
MINING_DATA_BUILD_V Training data for data mining (not text mining)
MINING_DATA_TEST_V Test data for data mining (not text mining)
MARKET_BASKET_V Data for association rules
MINING_DATA_ONE_CLASS_V Data for one-class SVM
You can see the references to tables in SH by listing the view definitions. The definition
of the view MINING_DATA_BUILD_V is shown as follows.
SQL> set long 1000000
SQL> set longc 100000
SQL> set pagesize 100
SQL> SELECT text FROM all_views WHERE
owner='dmuser3'AND view_name='mining_data_build_v';

The Sample Data
SELECT a.CUST_ID, a.CUST_GENDER, 2003-a.CUST_YEAR_OF_BIRTH AGE,

a.CUST_MARITAL_STATUS, c.COUNTRY_NAME, a.CUST_INCOME_LEVEL,
b.EDUCATION, b.OCCUPATION, b.HOUSEHOLD_SIZE, b.YRS_RESIDENCE,
b.AFFINITY_CARD, b.BULK_PACK_DISKETTES, b.FLAT_PANEL_MONITOR,
b.HOME_THEATER_PACKAGE, b.BOOKKEEPING_APPLICATION,
b.PRINTER_SUPPLIES, b.Y_BOX_GAMES, b.OS_DOC_SET_KANJI
FROM sh.customers a,
sh.supplementary_demographics b,
sh.countries c
WHERE a.CUST_ID = b.CUST_ID AND a.country_id = c.country_id
AND a.cust_id between 101501 and 103000
The views are used to build, test, and score the sample models. Each view has a
CUSTOMER_ID column, which is the case ID, and an AFFINITY_CARD column, which is
the target used by the predictive models. Most of the views provide data for 1500
customers (1500 rows). The view used by the One-Class SVM model has data for 940
customers.
The columns of training data in the MINING_DATA_BUILD_V view are as follows.
SQL> DESCRIBE mining_data_build_v
CUST_ID NOT NULL NUMBER

CUST_GENDER NOT NULL CHAR(1)
AGE NUMBER
CUST_MARITAL_STATUS VARCHAR2(20)
COUNTRY_NAME NOT NULL VARCHAR2(40)
CUST_INCOME_LEVEL VARCHAR2(30)
EDUCATION VARCHAR2(21)
OCCUPATION VARCHAR2(21)
HOUSEHOLD_SIZE VARCHAR2(21)
YRS_RESIDENCE NUMBER
AFFINITY_CARD NUMBER(10)
BULK_PACK_DISKETTES NUMBER(10)
FLAT_PANEL_MONITOR NUMBER(10)
HOME_THEATER_PACKAGE NUMBER(10)
BOOKKEEPING_APPLICATION NUMBER(10)
PRINTER_SUPPLIES NUMBER(10)
Y_BOX_GAMES NUMBER(10)
OS_DOC_SET_KANJI NUMBER(10)
Market Basket Data for Association Rules

The association demos use the MARKET_BASKET_V data set, which includes columns of
products from the PRODUCTS table and the CUSTOMER_ID column from the CUSTOMERS
table in SH. The columns of the MARKET_BASKET_V view are as follows.
SQL> DESCRIBE market_basket_v

EXTENSION_CABLE NUMBER
FLAT_PANEL_MONITOR NUMBER
CD_RW_HIGH_SPEED_5_PACK NUMBER
ENVOY_256MB_40GB NUMBER
ENVOY_AMBASSADOR NUMBER
EXTERNAL_8X_CD_ROM NUMBER
KEYBOARD_WRIST_REST NUMBER
SM26273_BLACK_INK_CARTRIDGE NUMBER
MOUSE_PAD NUMBER
MULTIMEDIA_SPEAKERS_3INCH NUMBER
OS_DOC_SET_ENGLISH NUMBER

The Sample Data
SIMM_16MB_PCMCIAII_CARD NUMBER
STANDARD_MOUSE NUMBER
Customer Data for Text Mining

The text mining demos use the same customer data from tables in SH, but they include
either an extra text column or a collection type column. The collection type is a nested
table of type DM_NESTED_NUMERICALS.
Tip: The process of extracting terms from a text column into a nested
table column is described in Oracle Data Mining User's Guide.
You can list these tables with the following SQL statements.
SQL>CONNECT dmuser
SQL>SELECT table_name FROM user_tables WHERE table_name LIKE '%MINING%';
The text mining tables are listed in Table 7–6.
Table 7–6 Tables Used by the Text Mining Sample Programs

Table Name Description
MINING_APPLY_NESTED_TEXT Apply table with COMMENTS column as DM_NESTED_NUMERICALS
MINING_BUILD_NESTED_TEXT Build table with COMMENTS column as DM_NESTED_NUMERICALS
MINING_TEST_NESTED_TEXT Test table with COMMENTS column as DM_NESTED_NUMERICALS
MINING_APPLY_TEXT Apply table with COMMENTS column as VARCHAR2(4000)
MINING_BUILD_TEXT Build table with COMMENTS column as VARCHAR2(4000)
MINING_TEST_TEXT Test table with COMMENTS column as VARCHAR2(4000)
In the MINING_BUILD_TEXT, MINING_TEST_TEXT, and MINING_APPLY_TEXT tables, the

COMMENTS column is of type VARCHAR2(4000).
SQL> DESCRIBE mining_build_text
Name Null? Type
----------------------------------------- -------- ----------------------------
AGE NUMBER
COMMENTS VARCHAR2(4000)

The Sample Data
In the MINING_*_NESTED_TEXT tables, the COMMENTS column is of type DM_NESTED_

NUMERICALS.
SQL> DESCRIBE mining_build_nested_text
Name Null? Type
----------------------------------------- -------- ----------------------------
AGE NUMBER
COMMENTS DM_NESTED_NUMERICALS

A
Verifying Your Data Mining Installation
A
This appendix provides a set of SQL and PL/SQL commands that test whether or not
the basic capabilities of data mining are present in the database. You can execute these
commands directly in the database without downloading or installing any additional
software.
Note: These commands create and score several data mining models.
These models do not perform meaningful data mining. They simply
show whether or not model creation and scoring operations are
working properly.
This appendix includes the following topics:

■ Connect to the Database
■ Create a Script or Execute the Commands Individually
■ Command Output
Connect to the Database

In SQL*Plus or SQL Developer, connect to the database as the system user. If you are
not sure how to start these tools, see "Connect Application Development Tools" on
page 3-1.
CONNECT sys / as sysdba
Enter password: <sys_password>
Insure that the USERS and TEMP tablespaces are present in the database. These
tablespaces are typically included by default and are always included in the starter
database created during installation (See "Install Oracle Database" on page 2-1). You
can check with a simple query.
select tablespace_name from dba_tablespaces;
Create a Script or Execute the Commands Individually

You can create a script by copying the commands in Example A–1 into a text file,
saving the file with a .SQL extension, and invoking the script from the SQL command
line.
SQL> @ script_name.SQL
Verifying Your Data Mining Installation A-1

Create a Script or Execute the Commands Individually
Alternatively, you can copy each command from Example A–1 to the SQL command
line and execute it individually.
Example A–1 Commands To Verify Data Mining Installation

--------------------------------------------------------
-- create a data mining user
--------------------------------------------------------
CREATE USER dm IDENTIFIED BY <dm_password>
DEFAULT TABLESPACE users
TEMPORARY TABLESPACE temp
QUOTA 20M ON users;
-- grant minimum privileges

GRANT create mining model TO dm;
GRANT create table TO dm;
GRANT create view TO dm;
GRANT create session TO dm;
--------------------------------------------------------
-- create a simple table for mining
--------------------------------------------------------
CONNECT dm/dm;
CREATE TABLE mt (p NUMBER, t NUMBER);
INSERT INTO mt VALUES (1,1);
INSERT INTO mt VALUES (2,2);
COMMIT;
--------------------------------------------------------
-- create models
--------------------------------------------------------
EXEC dbms_data_mining.create_model('MAI','ATTRIBUTE_IMPORTANCE','MT',null,'T');
EXEC dbms_data_mining.create_model('MAR','ASSOCIATION','MT','T');
EXEC dbms_data_mining.create_model('MNB','CLASSIFICATION','MT',null,'T');
EXEC dbms_data_mining.create_model('MSVM','REGRESSION','MT',null,'T');
EXEC dbms_data_mining.create_model('MKM','CLUSTERING','MT',null);
EXEC dbms_data_mining.create_model('MNMF','FEATURE_EXTRACTION','MT',null);
-- check the models

SELECT model_name, algorithm FROM user_mining_models;
--------------------------------------------------------
-- score models
--------------------------------------------------------
SELECT p,t,PREDICTION(MNB USING *) pred_nb,
p,t,prediction(MSVM USING *) pred_svm,
p,t,cluster_id(MKM USING *) clus,
p,t,feature_id(MNMF USING *) feat
FROM mt;
--------------------------------------------------------
-- drop the models and data
--------------------------------------------------------
EXEC dbms_data_mining.drop_model('MAI');
EXEC dbms_data_mining.drop_model('MAR');
EXEC dbms_data_mining.drop_model('MNB');
EXEC dbms_data_mining.drop_model('MSVM');
EXEC dbms_data_mining.drop_model('MKM');
EXEC dbms_data_mining.drop_model('MNMF');
A-2 Oracle Data Mining Administrator's Guide

Command Output
DROP TABLE mt PURGE;
Command Output
The output of the commands in Example A–1 is as follows.
User created.
Grant succeeded.
Grant succeeded.
Grant succeeded.
Grant succeeded.
Connected.
Table created.
1 row created.
1 row created.
Commit complete.
PL/SQL procedure successfully completed.
MODEL_NAME ALGORITHM
------------------------------ ------------------------------
MAI MINIMUM_DESCRIPTION_LENGTH
MAR APRIORI_ASSOCIATION_RULES
MKM KMEANS
MNB NAIVE_BAYES
MNMF NONNEGATIVE_MATRIX_FACTOR
MSVM SUPPORT_VECTOR_MACHINES
6 rows selected.
P T PRED_NB P T PRED_SVM P
---------- ---------- ---------- ---------- ---------- ---------- ----------
T CLUS P T FEAT
---------- ---------- ---------- ---------- ----------
1 1 1 1 1 1.48008198 1
1 3 1 1 1
2 2 2 2 2 1.51991802 2
2 2 2 2 1
Verifying Your Data Mining Installation A-3

Command Output
Table dropped.
A-4 Oracle Data Mining Administrator's Guide

Index
A enable or disable, 2-9

data mining user, 2-1, 2-11, 3-1, 4-1
Adaptive Bayes Network Database Companion
See deprecated features See Database Examples
ALL_MINING_MODEL_ATTRIBUTES, xi, 5-1 Database Configuration Assistant, xi, 1-2, 7-2
ALL_MINING_MODEL_SETTINGS, xi, 5-1, 5-3 Database Examples
ALL_MINING_MODELS, xi, 5-1 installation, 2-10
ALL_OBJECTS, xi sample programs, 7-1
ALTER ANY MINING MODEL, 4-3 database file location, 2-5
ALTER MINING MODEL, 4-4 database tuning, vi
anomaly detection, 7-7 Database Upgrade Assistant, 6-2
Apriori, 7-6, 7-9 DBA
association, 7-7, 7-9, 7-10, 7-12 database tuning, vi
attribute importance, 7-6, 7-8, 7-9, 7-10 production database, v, 2-1
attributes, 5-2 DBA_AUDIT_OBJECT, 5-6
AUDIT, 4-3, 5-4 to 5-6 DBA_MINING_MODEL_ATTRIBUTES, 5-1
AUDIT_TRAIL parameter, 5-4 DBA_MINING_MODEL_SETTINGS, 5-1
Automatic Data Preparation, 7-7 to 7-8 DBA_MINING_MODELS, 5-1, 6-3
DBA_OBJ_AUDIT_OPTS, 5-6
B DBA_OBJECTS, xi
DBA_REGISTRY, 6-3
binning, 7-9, 7-10
DBMS_DATA_MINING, 5-2, 7-6
BLAST
DBMS_DATA_MINING_TRANSFORM, 7-6
See desupported features
DBMS_DATA_MINING.DROP_MODEL, 6-4
Decision Tree, 7-6, 7-9
C demo programs
catalog views, xi, 5-1 See sample programs
chopt utility, 2-9 deprecated features, ix, xi, 1, 8
classification, 7-6, 7-7, 7-9, 7-10 desupported features, xi
CLASSPATH, 7-5 DHCP, 2-2
clipping, 7-9, 7-10 directory objects, 5-7
clustering, 7-7, 7-8, 7-9, 7-10 DM$P_MODEL_EXPIMP_TEMP table, 5-8
COMMENT, 4-3, 5-2, 5-5 DM$P_MODEL_IMPORT_TEMP table, 5-8
COMMENT ANY MINING MODEL, 4-3 DM$P_MODEL_TABKEY_TEMP table, 5-8
COMMENT ANY MODEL, 5-4 DM_USER_MODELS, xi
COMPATIBLE parameter, 6-3 DMEIDMSYS, 6-3
connection name, 3-2 dmshgrants.sql, 1-4, 5-10, 7-2
cost matrix, 4-4 dmsh.sql, 7-2, 7-11
CREATE ANY DIRECTORY, 5-7 DMSYS, x, xi, 6-3
CREATE ANY MINING MODEL, 4-3 downgrading, 6-1, 6-4
CREATE MINING MODEL, x, 1-2, 4-2, 4-3 DROP ANY MINING MODEL, 4-3
DSS, v
D
E
data dictionary views, xi, 5-1
Data Mining option, 2-1 Enterprise Edition, 2-1, 2-5
Index-1
EXPDP, 5-6 adding a comment, 5-4
EXPLAIN, 2-11, 2-13, 7-10 applying, 4-4
exporting, 2-13, 4-2, 5-6 to 5-8, 6-2, 7-9, 7-10 auditing, 5-4
changing the name, 4-4
listing, 5-1
F object privileges, 4-4
feature extraction, 7-7, 7-8, 7-9, 7-10 See also scoring
settings, 5-3
G upgrading, 6-1
viewing, 4-4
Generalized Linear Models, 7-6, 7-7, 7-10 model details, 5-2
GET_DEFAULT_SETTINGS, xi model signature, xi, 5-2
GET_MODEL_DETAILS, 5-2
GET_MODEL_SETTINGS, xi
GET_MODEL_SIGNATURE, xi N
GLM Naive Bayes, 7-7, 7-9
See Generalized Linear Models Net Configuration Assistant, 2-11, 3-8
global database name, 2-5, 3-3 Net Service Name, 2-12, 3-8
new features, ix
I NMF
See Non-Negative Matrix Factorization
IMPDP, 5-6 Non-Negative Matrix Factorization, 7-7, 7-8, 7-9,
importing, 2-13, 4-2, 5-6 to 5-9, 6-2, 7-9, 7-10 7-11
INIT.ORA, vi normalization, 7-9, 7-10
installation
Data Mining sample programs, 7-1
Oracle Client, 3-4 O
Oracle Data Miner, 2-11 object privileges, 4-4
Oracle Database, 2-1 O-Cluster, 7-7, 7-9
Oracle Database Examples, 2-10 OLTP, v
Oracle Spreadsheet Add-In for Predictive One-Class SVM, 7-7
Analytics, 2-11 Oracle base directory, 2-5, 2-11
Oracle By Example, 3
J Oracle Client
connect, 3-3
Java API, ix, 6-1, 1, 8 install, 3-4
See also deprecated features required for Spreadsheet Add-In, 2-11
Java API sample programs, 7-8 Oracle Data Miner, ix, 3, 3-1, 2, 3-3, 4
Java API text mining, 7-10 Oracle Data Pump, 5-6
JDM 1.0 standard, 7-8 Oracle Database
JSR-73 specification, 7-8 Enterprise Edition, 2-1, 2-5
install, 2-1
K local, 3-2
remote, 3-3
k-Means, 7-6, 7-9 Oracle home directory, 2-5, 2-10, 2-11, 2-13, 3-6, 7-2,
7-4, 7-5
L Oracle Net, 3-3
Oracle Objects for OLE, 2-11
linear regression, 7-6, 7-7, 7-9
Oracle services, 2-2
logistic regression, 7-6, 7-7, 7-9
Oracle Spreadsheet Add-In for Predictive Analytics
loopback adapter, 2-2
install, 2-12
Net Service Name, 3-3, 3-8
M start, 2-12
market basket data, 7-12 Oracle Text, 7-7
memory tuning parameters, vi ORACLE_BASE, 2-5
Microsoft Excel, 1-4, 2-11, 2-12, 3-12 ORACLE_HOME, vii
Microsoft Windows, 2-1
Minimum Description Length, 7-6, 7-9 P
mining models
parallel execution, vi
about, 5-1
performance, vi
Index-2
PGA_AGGREGATE_TARGET, vi See Also One-Class SVM
PL/SQL API, 3-1, 6-1 See Support Vector Machine
PL/SQL sample programs, 7-6 SYS, x, 2-5
PL/SQL text mining, 7-8 system privileges, 4-3
PMML, ix, 5-11
port number, 3-2
T
PREDICT, 2-11, 2-13, 7-10
predictive analytics, 2-11, 3-3, 7-8, 7-9, 7-10 term extraction, 7-7, 7-8
privileges for data mining, x, 2-1, 3-1, 4-1, 5-7, 7-1 text mining, 7-7, 7-8, 7-9, 7-10, 7-13
privileges for exporting and importing, 5-7 transformations, 7-9
privileges for running sample programs, 7-1, 7-2 tuning, vi
production database, v, 2-1
PROFILE, 2-11, 2-13, 7-10 U
upgrading, 6-1
R USER_AUDIT_OBJECT, 5-6
regression, 7-6, 7-7, 7-9, 7-10 USER_MINING_MODEL_ATTRIBUTES, 5-1
USER_MINING_MODEL_SETTINGS, 5-1
USER_MINING_MODELS, 5-1
S USER_OBJ_AUDIT_OPTS, 5-6
sample mining models, 7-5 USER_OBJECTS, xi
sample programs, 2-10, 7-1 users, 2-1, 3-2, 5-7, 7-1, 7-2
configuration scripts, 4-3, 7-1, 7-2
data used by, 7-11
find, 7-2
install, 7-1
Java, 7-4, 7-8
PL/SQL, 7-6
run, 7-4, 7-5
user, 4-3, 7-1
sample schemas, 7-1
scoring
batch, v, vi
java programs, 7-9
model signature, 5-2
privileges for, 4-3
real-time, vi
single-row, v
Scoring Engine, 6-2
See also desupported features
security, x, 4-1
SELECT ANY MINING MODEL, 4-3
SELECT MINING MODEL, 4-4
SGA_TARGET, vi
SH schema, 7-1, 7-2, 7-11
shared cursor, vi
SID, 3-3
software location
See Oracle home directory
Spreadsheet Add-In
See Oracle Spreadsheet Add-In for Predictive
Analytics
start, 2-12
SQL AUDIT, 5-4
SQL COMMENT, 5-4
SQL Developer, 3-1, 3-2, 4
SQL*Plus, 3-1
starter database., 2-5
Support Vector Machine, 7-7, 7-8, 7-9, 7-11
SVM
Index-3
Index-4

Data Mining INST

Uploaded by

Copyright:

Available Formats

Data Mining INST

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Data Mining INST

Uploaded by

Copyright:

Available Formats

Oracle® Data Mining

Primary Author: Kathy L. Taylor

What's New in Oracle Data Mining Administration? ............................................................... ix

1 Getting Started with Oracle Data Mining

2 Installing Oracle Data Mining

3 Connecting to an Oracle Database

4 Users and Privileges for Data Mining

5 Working With Mining Model Schema Objects

6 Upgrading Oracle Data Mining

7 The Data Mining Sample Programs

A Verifying Your Data Mining Installation

Note for DBAs: DBAs managing production databases that support

Access to Oracle Support

Where to Find Database Tuning Documentation

See Also: Chapter "Using Parallel Execution" in Oracle Database

This section summarizes the new features in Oracle Data Mining.

Oracle Database 11g Release 2 (11.2.0.3) Oracle Data Mining

Note: Oracle recommends that you not use deprecated features in

Oracle Database 11g Release 2 (11.2.0.2) New Features in Oracle Data

Oracle Database 11g Release 1 New Features in Oracle Data Mining

New Database Privilege Required for Creating Models

See Also: Chapter 4, "Users and Privileges for Data Mining"

Mining Models in the Oracle Data Dictionary

See Also: "Obtaining Information from the Data Dictionary" in

Features Not Available in Oracle Data Mining 11g Release 1

Features Deprecated in Oracle Data Mining 11g Release 1

Note: Oracle recommends that you do not use deprecated features in

Note: The minimum requirements for a working installation of

■ Step One: Install the Database

Step One: Install the Database

Getting Started with Oracle Data Mining 1-1

3. The Installer performs prerequisite checks, displays summary information about

See: "Install Oracle Database" on page 2-1 for more detailed

Step Two: Verify the Data Mining Installation

See: Appendix A, "Verifying Your Data Mining Installation"

Step Three: Create a Data Mining User

3. Execute these statements to grant data mining privileges to the user.

GRANT SELECT ON data TO dmuser; -- required for mining data that is

1-2 Oracle Data Mining Administrator's Guide

Step Four: If You Want to Use Oracle Data Miner

Step Five: If You Want to Use the Spreadsheet Add-In

Getting Started with Oracle Data Mining 1-3

To install and launch the Spreadsheet Add-In:

See: "Optionally Install the Spreadsheet Add–In" on page 2-11 for

Step Six: If You Want to Use the Sample Programs

See: Chapter 7, "The Data Mining Sample Programs" for more

1-4 Oracle Data Mining Administrator's Guide

Note: For production installations of Oracle Database with Oracle

This chapter contains the following sections:

Install Oracle Database

Installing Oracle Data Mining 2-1

Important: These instructions are not intended as a replacement for

Note: If your computer uses Dynamic Host Configuration Protocol

2. Log on to your computer as a member of the Administrator's group. If you are

Note: To stop an Oracle database, first stop the database service

2-2 Oracle Data Mining Administrator's Guide

Installing Oracle Data Mining 2-3

Note: If you choose Install database software only instead of Create

7. On the System Class page, choose Desktop Class then Next.