DW OLAP DM
,,, :, DW OLAP DM DW OLAP DM
DW OLAP DM
80 " "William H.Inmon
Data mart " "
Meta Data
( )
ETL ETL Extract Transform Cleansing Load
ETL
Oracle Oracle (CWM) (OEM) 9i Application Server 9i Application Server BI Beans BI Beans Discoverer Discoverer Reports Reports Portal Portal Oracle9i ETL OLAP Oracle9i ETL OLAP Warehouse Builder Warehouse Builder
Oracle Warehouse Builder
OLAP
OLAP? OLAP (OLAP) E.F.Codd 1993 Codd (OLTP) SQL Codd OLAP
OLAP 1 OLAP 2 3 4 5 / 6 7 8 9 10 11 12
OLAP OLAP ROLAP MOLAP HOLAP Server OLAP Client OLAP
Client OLAP Client OLAP Server OLAP Brio Designer Business Object
ROLAP ROLAP Codd 12 star schema OLAP Relational OLAP ROLAP Informix Metacube Microsoft SQL Server OLAP Services
ROLAP (Star Schema) Period Table Time_id Period_Desc Quarter Year "Dimension Table" Product Table Product_id Product_Desc Brand Size "Dimension Table" Sales Table Time_id Product_id Market_id Scenario Dollars Units Discount% "Fact Table" Market Table Market_id Market_Desc District Region "Dimension Table" Scenario Table Scenario Actual Profit Budget "Dimension Table"
MOLAP MOLAP Arbor Software Codd Muiltdimension OLAP MOLAP Hyperion( Arbor Software) Essbase Showcase Strategy
MOLAP (Multicube)
OLTP ROLAP MOLAP
RDB - MDDB 50 60 100 40 70 80 90 120 140 50 60 100 40 70 80 90 120 140
OLAP 1. ( ) 2. ( ) ( ) 3. ( ) 4. ( 1 2 n ) ( ) 5. ( ) (2000 1 $100000)
OLAP 1. (Slice and Dice),,, 2. (Drill) (Drill-down) up)/ (Roll-up) 3. (Rotate)/ (Pivot)
OLAP ( )
OLAP ( ) 60
OLAP ( )
OLAP (1) : OLAP 5 (2) :OLAP (3) : OLAP, (4) : OLAP
Oracle Express
DM AI (Knowledge Discover Database KDD)
K-
7
Cases Historic Data Attributes Name Income Age....... Jones 30,000 30 Smith 55,000 67 Lee 25,000 23 Rogers 50,000 44 Buy my Product? 1 =Yes, 0 =No 1 1 0 0 Functional Relationship: Model Y = F(X 1, X 2,, X m ) X 1 X 2... Xm Y Predictive Attributes Supplemental Attribute Target Attribute
Cases Historic Data Attributes Name Income Age....... Jones 30,000 30 Smith 55,000 67 Lee 25,000 23 Rogers 50,000 44 Campos40,500 Hornick 37,000 Habers 57,200 Berger 95,600 New Records 52 73 32 34 Buy my Product? 1 =Yes, 0 =No Predictions 1 1 0 0? 1? 0? 0? 1.85.74.93.65 Functional Relationship: Model Y = F(X 1, X 2,, X m ) Confidence
7
The Data Mining Process Problem Definition Business Domain Expert Data Preparation DBA Data Analyst Model Building & Testing Data Analyst Java Developer Deploy & Apply Model Java Developer End Users
Oracle9i Complete e-business Intelligence Oracle9iDB Data Warehousing Oracle9iAS Portal ETL OLAP Data Mining M e t a d a t a Query & Reporting BI Components Real-Time Personalization Hello! We have recommendations for you.
Oracle11i CRM Application CRM / Data Mining Integration Marketing analysts can design targeted campaigns without becoming data mining experts Build models, score lists Discover patters & make predictions Data mining increases effectiveness of targeted campaigns
Oracle11i CRM Application Marketing analysts can simply select the top deciles for more targeted campaigns CRM / Data Mining Integration Target customers most likely to respond