Contents

Title and Copyright Information

Preface

• Audience
• Documentation Accessibility
• Related Documentation
• Conventions

1 Introducing the Oracle Data Mining APIs

• 1.1 New Features
• 1.2 Predictive and Descriptive Data Mining
• 1.3 Steps in a Data Mining Application
• 1.4 Data Preparation
• 1.5 Model Settings
• 1.6 Model Details
• 1.7 Predictive Analytics
• 1.8 SQL Scoring Functions

2 Managing Data

• 2.1 Data Types
• • 2.1.1 Collection Types
  • 2.1.2 Text
  • 2.1.3 Date and Time Data
• 2.2 Columns and Attributes
• • 2.2.1 Attribute Data Types
  • • 2.2.1.1 Converting Column Data Types
    • 2.2.1.2 DATE and TIMESTAMP Columns with Predictive Analytics
  • 2.2.2 Attribute Names
• 2.3 Nested Tables
• • 2.3.1 Object Views and Multi-Record Collections
  • 2.3.2 Example: Multi-Record Collections With an Object View
• 2.4 Data Storage Optimization

3 Managing Models

• 3.1 Models in the Database
• • 3.1.1 Model Names
  • 3.1.2 Model Access
• 3.2 Import/Export
• 3.3 Model Settings
• • 3.3.1 Costs
  • 3.3.2 Priors

4 Using the PL/SQL API and SQL Scoring Functions

• 4.1 The PL/SQL Sample Applications
• 4.2 The DBMS_DATA_MINING Package
• • 4.2.1 Build Results
  • 4.2.2 Apply Results
  • 4.2.3 Test Results for Classification Models
  • 4.2.4 Test Results for Regression Models
  • • 4.2.4.1 Root Mean Square Error
    • 4.2.4.2 Mean Absolute Error
• 4.3 Example: Building a Decision Tree Model
• • 4.3.1 Mining Data
  • 4.3.2 Build Settings
  • 4.3.3 Model Creation
• 4.4 Example: Using SQL Functions to Test a Decision Tree Model
• 4.5 Example: Using SQL Functions to Apply a Decision Tree Model

5 Using PL/SQL to Prepare Text Data for Mining

• 5.1 Oracle Text for Oracle Data Mining
• 5.2 Term Extraction in the Sample Programs
• • 5.2.1 Text Mining Programs
• 5.3 From Unstructured Data to Structured Data
• 5.4 Steps in the Term Extraction Process
• • 5.4.1 Transform a Text Column in the Build Table
  • 5.4.2 Transform a Text Column in the Test and Apply Tables
  • 5.4.3 Creating the Index and Index Preference
  • 5.4.4 Creating the Intermediate Terms Table
  • • 5.4.4.1 FEATURE_PREP Calling Syntax
    • 5.4.4.2 FEATURE_PREP Return Value
    • 5.4.4.3 FEATURE_PREP Arguments
    • 5.4.4.4 FEATURE_PREP Example
  • 5.4.5 Creating the Final Terms Table
  • • 5.4.5.1 FEATURE_EXPLAIN Calling Syntax
    • 5.4.5.2 FEATURE_EXPLAIN Return Value
    • 5.4.5.3 FEATURE_EXPLAIN Arguments
    • 5.4.5.4 FEATURE_EXPLAIN Example
  • 5.4.6 Populating a Nested Table Column
• 5.5 Example: Transforming a Text Column

6 Java API Overview

• 6.1 The JDM 1.0 Standard
• 6.2 Oracle Extensions to JDM 1.0
• 6.3 Principal Objects in the ODM Java API
• • 6.3.1 PhysicalDataSet Object
  • 6.3.2 BuildSettings Object
  • 6.3.3 Task Object
  • 6.3.4 Model Object
  • 6.3.5 TestMetrics Object
  • 6.3.6 ApplySettings Object

7 Using the Java API

• 7.1 The Java Sample Applications
• 7.2 Setting up Your Development Environment
• 7.3 Connecting to the Data Mining Server
• • 7.3.1 Connection Factory
  • • 7.3.1.1 Create a ConnectionFactory Using OraConnectionFactory
    • 7.3.1.2 Create a ConnectionFactory From the JNDI Server
  • 7.3.2 Managing the DMS Connection
  • • 7.3.2.1 Pre-Create the JDBC Connection
    • 7.3.2.2 Use a ConnectionSpec for the DMS Connection
  • 7.3.3 Features of a DMS Connection
  • • 7.3.3.1 Create Object Factories
    • 7.3.3.2 Provide Access to Mining Object Metadata
    • 7.3.3.3 Save and Retrieve Mining Objects
    • 7.3.3.4 Execute Mining Tasks
    • 7.3.3.5 Retrieve DMS Capabilities and Metadata
    • 7.3.3.6 Retrieve Version Information
• 7.4 API Design Overview
• 7.5 Describing the Mining Data
• 7.6 Build Settings
• 7.7 Executing Mining Tasks
• 7.8 Building a Mining Model
• 7.9 Exploring Model Details
• 7.10 Testing a Model
• 7.11 Applying a Model for Scoring Data
• 7.12 Using a Cost Matrix
• 7.13 Using Prior Probabilities
• 7.14 Using Automated Prediction and Explain Tasks
• 7.15 Preparing the Data
• • 7.15.1 Using Binning/Discretization Transformation
  • 7.15.2 Using Normalization Transformation
  • 7.15.3 Using Clipping Transformation
  • 7.15.4 Using Text Transformation

8 Converting to the ODM 10.2 Java API

• 8.1 Comparing the 10.1 and 10.2 Java APIs
• 8.2 Converting Your Applications

9 Sequence Matching and Annotation (BLAST)

• 9.1 NCBI BLAST
• 9.2 Using ODM BLAST
• • 9.2.1 Using BLASTN_MATCH to Search DNA Sequences
  • • 9.2.1.1 Searching for Good Matches in DNA Sequences
    • 9.2.1.2 Searching DNA Sequences Published After a Certain Date
  • 9.2.2 Using BLASTP_MATCH to Search Protein Sequences
  • • 9.2.2.1 Searching for Good Matches in Protein Sequences
  • 9.2.3 Using BLASTN_ALIGN to Search and Align DNA Sequences
  • • 9.2.3.1 Searching and Aligning for Good Matches in DNA Sequences
  • 9.2.4 Output of BLAST Queries
  • 9.2.5 Using BLASTN_COMPRESS to Improve Search Performance
  • • 9.2.5.1 Compress Sequences
    • 9.2.5.2 Passing a Compressed Sequence to a BLAST Function
  • 9.2.6 Sample Data for BLAST
  • • 9.2.6.1 SWISSPROT Table
    • 9.2.6.2 PROT_DB Table
    • 9.2.6.3 ECOLI10 Table
    • 9.2.6.4 Genetic Codes and Names
    • 9.2.6.5 Sequence Databases
    • 9.2.6.6 Loading Sequences into an Oracle Database
• Summary of BLAST Table Functions
• • BLASTN_COMPRESS Table Function
  • BLASTN_MATCH Table Function
  • BLASTP_MATCH Table Function
  • TBLAST_MATCH Table Function
  • BLASTN_ALIGN Table Function
  • BLASTP_ALIGN Table Function
  • TBLAST_ALIGN Table Function

Index