Class15 - Data Warehousing

Data Warehousing
 Data
 Raw piece of information that is capable of being moved
and store.
 Database
 An organized collection of such data in which data are
managed in tabular form with relationship.
 Data Warehouse
 System that organizes all the data available in an
organization, makes it accessible & usable for the all kinds
of data analysis and also allows to create a lots of reports
by the use of mining tools.
CH#2, Data Warehousing By: Babu Ram Dawadi
Data Warehouse…
 “A data warehouse is a subject-oriented,
integrated, time-variant, and nonvolatile collection
of data in support of management’s decision-
making process.”
 Data warehousing:
 The process of constructing and using data
warehouses.
 Is the process of extracting & transferring
operational data into informational data & loading
it into a central data store (warehouse)
Data Warehouse—Integrated
 Constructed by integrating multiple,
heterogeneous data sources Sales
 relational databases, flat files, on-line system
transaction records
 Data cleaning and data integration
techniques are applied. Payroll
system Customer
 Ensure consistency in naming
data
conventions, encoding structures,
attribute measures, etc. among different
data sources Purchasing
 E.g., Hotel price: currency, tax, system
breakfast covered, etc.
 When data is moved to the warehouse, it
is converted.
Data Warehouse—Subject-
Oriented
 Organized around major subjects, such
as customer, product, sales.
Sales Employee
 Focusing on the modeling and analysis system data
of data for decision makers, not on daily

operations or transaction processing. Payroll Customer
system data
 Provide a simple and concise view
around particular subject issues by Vendor
Purchasing data
excluding data that are not useful in the system
decision support process. Operational data DW

Data Warehouse—Time Variant
 The time horizon for the data warehouse is
significantly longer than that of operational systems.
 Operational database: current value data.
 Data warehouse data: provide information from a historical
perspective (e.g., past 5-10 years)
 Every key structure in the data warehouse
 Contains an element of time, explicitly or implicitly
 But the key of operational data may or may not contain
“time element”.
Data Warehouse—Non-Volatile
 A physically separate store of data DBMS DW
transformed from the operational
create access
environment.
 Operational update of data does not occur
in the data warehouse environment. update Customer
Sales delete
 Does not require transaction processing, system data
recovery, and concurrency control
mechanisms
insert load
 Requires only two operations in data
accessing:
 initial loading of data and access of data.

Data Warehouse Usage
 Three kinds of data warehouse applications
 Information processing
supports querying, basic statistical analysis, and reporting
using crosstabs, tables, charts and graphs
 Analytical processing
 multidimensional analysis of data warehouse data
 supports basic OLAP operations, slice-dice, drilling,
pivoting
 Data mining
 knowledge discovery from hidden patterns
 supports associations, constructing analytical models,
performing classification and prediction, and presenting
the mining results using visualization tools.
The Warehousing Approach
 Information Clients
integrated in
advance Data
Warehouse
 Stored in WH
for direct
querying and Integration System Metadata
analysis
...
Extractor/ Extractor/ Extractor/
Monitor Monitor Monitor
CH#2, Data Warehousing

...
By: Babu Ram Dawadi
Source Source Source
General Architecture
OLAP
External Data Server Data
Sources
acquisition extraction OLAP
queries/
Query reports
Data and
Integration Data Data Analysis
Component Warehouse Component
data
mining
Metadata
Internal Monitoring
Sources Administration
Construction &
maintenance
3 main phases
 Data acquisition
 relevant data collection
 Recovering: transformation into the data warehouse model from
existing models
 Loading: cleaning and loading in the DWH
 Storage
 Data extraction
 Tool examples: Query report, SQL, multidimensional analysis
(OLAP tools), datamining
 + evolution and maintenance
10
DW Monitoring
 Identify growth factors and rate
 Identify what data is being used
 Identify who is using the data, and when
  Avoid constant growth

  Plan for evolution (trends)
 Control response time (latency)
11
DATA WAREHOUSING
THE USE OF A DATA WAREHOUSE
INVENTORY
DATABASE STEP 1: Load the Data Warehouse
PERSONNEL STEP 2: Question the Data Warehouse

DATABASE
DATA
NEWCASTLE
SALES DB WAREHOUSE
LONDON
SALES DB
STEP 3: Do something DECISIONS

and ACTIONS!
GLASGOW
with what you learn from
SALES DB the Data Warehouse
Partitioning
 To improve performances & flexibility without
giving up on the details
DW
 Data marts
 By date, business type, geography, …

13
The Need for Data Analysis
 Managers must be able to track daily
transactions to evaluate how the business is
performing
 By tapping into the operational database,

management can develop strategies to meet
organizational goals
 Data analysis can provide information about

short-term tactical evaluations and strategies

Creating a Data Warehouse

Factors Common to Data
Warehousing
 Dynamic framework for decision support
that is always a work in progress
 Must satisfy:
 Data integration and loading criteria
 Data analysis capabilities with acceptable
query performance
 End-user data analysis needs
 Apply database design procedures

Why Separate Data Warehouse?
 High performance for both systems
 DBMS— tuned for OLTP: access methods, indexing,
concurrency control, recovery
 Warehouse—tuned for OLAP: complex OLAP queries,
multidimensional view, consolidation(aggregation).
 Different functions and different data:
 missing data: Decision support requires historical data
which operational DBs do not typically maintain
 data consolidation: Decision Support requires
consolidation (aggregation, summarization) of data
from heterogeneous sources
 data quality: different sources typically use inconsistent
data representations, codes and formats
17
Decision Support Systems
 Methodology (or series of methodologies) designed to
extract information from data and to use such
information as a basis for decision making
 Decision support system (DSS):

 Arrangement of computerized tools used to assist managerial
decision making within a business
 Usually requires extensive data “massaging” to produce
information
 Used at all levels within an organization
 Often tailored to focus on specific business areas
 Provides ad hoc query tools to retrieve data and to display
data in different formats

Decision Support Systems
 Extract Information from data to use as the basis
for decision making
 Used at all levels of the Organization
 Tailored to specific business areas
 Interactive
 Ad Hoc queries to retrieve and display
information
 Combines historical operation data with
business activities
4 Components of DSS
 Data Store – The DSS Database
 Business Data
 Business Model Data
 Internal and External Data
 Data Extraction and Filtering
 Extract and validate data from the operational
database and the external data sources
4 Components of DSS
 End-User Query Tool
 Create Queries that access either the Operational
or the DSS database
 End User Presentation Tools
 Organize and Present the Data
Differences with DSS
 Operational
 Stored in Normalized Relational Database
 Support transactions that represent daily
operations (Not Query Friendly)
 3 Main Differences
 Time Span
 Granularity
 Dimensionality
Time Span
 Operational
 Real Time
 Current Transactions
 Short Time Frame
 Specific Data Facts
 DSS
 Historic
 Long Time Frame (Months/Quarters/Years)
 Patterns
Granularity
 Operational
 Specific Transactions that occur at a given time
 DSS
 Shown at different levels of aggregation
 Different Summary Levels
 Decompose (drill down)
 Summarize (roll up)
Dimensionality
 Most distinguishing characteristic of DSS
data
 Operational
 Represents atomic transactions
 DSS
 Data is related in Many ways
 Develop the larger picture
 Multi-dimensional view of data
DSS Database Requirements
 DSS Database Scheme
 Support Complex and Non-Normalized data
 Summarized and Aggregate data
 Multiple Relationships
 Queries must extract multi-dimensional time slices
 Redundant Data
 Data Extraction and Filtering
 DSS databases are created mainly by extracting data
from operational databases combined with data
imported from external source
 Need for advanced data extraction & filtering tools
 Allow batch / scheduled data extraction
 Support different types of data sources
 Check for inconsistent data / data validation rules
 Support advanced data integration / data formatting conflicts
 End User Analytical Interface
 Must support advanced data modeling and data
presentation tools
 Data analysis tools
 Query generation
 Must Allow the User to Navigate through the DSS
 Size Requirements
 VERY Large – Terabytes
 Advanced Hardware (Multiple processors, multiple disk
arrays, etc.)
Data Warehouse
 DSS – friendly data repository for the DSS is
the DATA WAREHOUSE
 Definition: Integrated, Subject-Oriented,

Time-Variant, Nonvolatile database that
provides support for decision making
Main Components of a
Decision Support System (DSS)

Transforming Operational Data
Into Decision Support Data

ON-LINE ANALYTICAL
PROCESSING (OLAP)

OLAP
WHAT IS OLAP?
DEFINITION :
‘OLAP applications and tools are those that are designed to ask
ad hoc, complex queries of large multidimensional collections
of data. It is for this reason that OLAP is often mentioned in the
context of Data Warehouses’.

The Multidimensional Idea
Region
Sales Granularity
Year Product
category
Quarter Product
type
Product
3 dimensions
34
OLAP
MULTDIMENSIONAL DATA MODEL
London
Glasgow Socks
Newcastle
Jumpers
10 50 10 10 T-Shirts
0 0 1 2 Shorts
80 80 80 80
Pyjamas
0 25 20 15
0 0 0 0
Spring Summer Autumn Winter
Example: Three dimensions By:

CH#2, Data Warehousing
– Product, Sales, Area, and Season
Babu Ram Dawadi
Storage: The Cube
Region
Mobiles Fax Standard
Vaud
Fribourg
Neuchatel
1999 Product
type
1998
1997
Year
Sales of standard telephones
in 1997 in Vaud region
36
OLAP Terminology
 A data cube supports viewing/modelling of a
variable (a set of variables) of interest. Measures
are used to report the values of the particular
variable with respect to a given set of dimensions.
 A fact table stores measures as well as keys

representing relationships to various dimensions.
 Dimensions are perspectives with respect to which

an organization wants to keep record.
 A star schema defines a fact table and its

associated dimensions. 37
3-D Cube
Fact table view: Multi-dimensional cube:

sale prodId storeId date amt
p1 c1 1 12
p2 c1 1 11 c1 c2 c3
p1 c3 1 50
day 2
p1 44 4
p2 c2 1 8 p2 c1 c2 c3
p1 c1 2 44 day 1
p1 12 50
p1 c2 2 4 p2 11 8
dimensions = 3

Typical OLAP Operations
 Roll up (drill-up): summarize data
 by climbing up hierarchy or by dimension reduction
 Drill down (roll down): reverse of roll-up
 from higher level summary to lower level summary or detailed
data, or introducing new dimensions
 Slice and dice:
 project and select
 Pivot (rotate):
 reorient the cube, visualization, 3D to series of 2D planes.
 Other operations
 drill across: involving (across) more than one fact table
Han: Dataware Houses and OLAP 39

OLAP
TYPICAL OLAP OPERATIONS
Drill Total Sales Drill

Down Total Sales per city Up
Total Sales per city per store
Total Sales per city per store per month
Drill Total Sales Drill

Down Total Sales per city Up
Total Sales per city by category
Drill Across

By a drill up opperation examine sales
By country rather than city level
location by city location y country

roll up
Istanbul Ankara Berlin Münih Türkiyy Almanya
PC 20 30 50 40 PC 50 90
Printer 15 5 10 20 Printer 20 30
measure is sales Drill down 2002

Time 2002 Q1 Q2 Q3 Q4
PC 10 15 20 5
PC 50 Printer 5 10 5 3
Printer 23

 when performed by dimension reduction
 one or more dimensions are removed from the cube
 Ex a sales cube with location and time
 roll-up may remove the time dimension
 aggregation of total sales by location
 rather than by location and by time
location by country
locat All
Türkiye Almanya
PC 50 90 PC 140
Printer 20 30 Printer 50
Two dimensional cuboid
One dim. cuboid 42
Roll-up and Drill-down algebraic operators
3
25 28
9
1997 180 244 72 35 180 244 72 44
9
42 51
5
318 204 78 22 318 204 78 27
1998
8
3 11
35 Lausane }Vaud
1999
131 153 57 11 Vevey 131 153 57 46 Vaud
Brig}Valais
Sion Valais
mobiles fax standard mobiles fax standard
IOION
N
WWN
Roll-up
N
EGG
TTOO
RRE
Less detailed: go up in the granularity hierarchy
Drill-down
More detailed: 43go down in the granularity hierarchy

Slice and dice
 Slice: a selection on one dimension of the cube
resulting in subcube
 Ex: sales data are selected for dimension time
using time =spring
 dice: defines a subcube by performing a selection
on two or more dimensions
 Ex: a dice opp. Based on
 location=“london” or “glasgow” and
 time =spring or summer and
 item = “T-shirts” or “Pyjamas”

N-DİMENSİONAL CUBE
 A data cube is referred to as a cuboid
 The lattice of cuboids forms a data cube.
 The cuboid holding the lowest level of summarization is

called a base cuboid.
 the 4-D cuboid is the base cuboid for the given four
dimensions
 The top most 0-D cuboid, which holds the highest-level of

summarization, is called the apex cuboid.
 typically denoted by all
45
Cube: A Lattice of Cuboids
all
0-D(apex) cuboid
time item location supplier

1-D cuboids
time,item time,location item,location location,supplier

2-D cuboids
time,supplier item,supplier
time,location,supplier
time,item,location 3-D cuboids
time,item,supplier item,location,supplier
46
4-D(base) cuboid
time, item, location, supplier
CONCEPTUAL MODELING OF
DATA WAREHOUSES
 Modeling data warehouses: dimensions & measures

 Starschema: A fact table in the middle connected to a set of
dimension tables
 Snowflake schema: A refinement of star schema where some
dimensional hierarchy is normalized into a set of smaller
dimension tables, forming a shape similar to snowflake
 Fact constellation: Multiple fact tables share dimension
tables, viewed as a collection of stars, therefore called galaxy
schema or fact constellation
47
EXAMPLE OF STAR SCHEMA
time
time_key item
day item_key
day_of_the_week Sales Fact Table item_name
month brand
quarter time_key type
year supplier_type
item_key
branch_key
branch location
location_key
branch_key location_key
branch_name units_sold street
branch_type city
dollars_sold province_or_street
country
avg_sales
48
Measures
STAR SCHEMA FOR SALES
Dimension
Tables
Fact Table
EXAMPLE OF SNOWFLAKE SCHEMA
time
time_key item
day item_key supplier
day_of_the_week Sales Fact Table item_name supplier_key
month brand supplier_type
quarter time_key type
year item_key supplier_key
branch_key
location
branch location_key
location_key
branch_key
units_sold street
branch_name
city_key city
branch_type
dollars_sold
city_key
avg_sales city
province_or_street
51
Measures country
EXAMPLE OF FACT
CONSTELLATION
time
time_key item Shipping Fact Table
day item_key
day_of_the_week Sales Fact Table item_name time_key
month brand
quarter time_key type item_key
year supplier_type shipper_key
item_key
branch_key from_location
branch location_key location to_location

branch_key location_key dollars_cost
branch_name units_sold
street
branch_type dollars_sold city units_shipped
province_or_street
avg_sales country shipper
Measures shipper_key
52
shipper_name
location_key
shipper_type
OLAP CLIENT/SERVER ARCHITECTURE
OLAP SERVER ARRANGEMENT
OLTP VS. OLAP
OLTP OLAP
users clerk, IT professional knowledge worker
function day to day operations decision support
DB design application-oriented subject-oriented
data current, up-to-date historical,
detailed, flat relational summarized, multidimensional
isolated integrated, consolidated
usage repetitive ad-hoc
access read/write lots of scans
index/hash on prim. key
unit of work short, simple transaction complex query
# records accessed tens millions
#users thousands hundreds
DB size 100MB-GB 100GB-TB
metric transaction throughput query throughput, response
55
OLTP –VS- OLAP
 On Line Transaction Processing -- OLTP
 Maintain a database that is an accurate model of some
real-world enterprise
 Short simple transactions
 Relatively frequent updates
 Transactions access only a small fraction of the database
 On Line Analytic Processing -- OLAP

 Use information in database to guide strategic decisions
 Complex queries
 Infrequent updates (Load)
 Transactions access a large fraction of the database

Business Information:
“How you gather, manage,
and use information
will determine whether
you win or lose.”
– Bill Gates

What is BI?
 Business Intelligence means using your data assets to make
better business decisions.
 Business intelligence involves the gathering, management, and

analysis of data for the purpose of turning that data into useful
information which is then used to improve decision making.
 Organizations can then make more strategic decisions about

how to administer clients and programs. These practices can
also reduce operating costs through more effective financial
analysis, risk management, and fraud management.

Business Intelligence solutions start with
data warehouses and data marts
Analysis Complexity & Value

Optimization
Discovery
Data Mining
Verification Multidimensional
Statistical
Data Mart
Data Warehouse
Stage 1 Stage 2 Stage 3 Stage 4 Stage 5

Data… Information….Decisions
Data to Information to Decisions
Data Information Access

Management
On-line Updates Data Warehouse Query & Reporting

Batch Feeds Data Mart Data Mining
Operational Data Store Data Transformation On-line Analytical
Processing
Data Synchronization
Summary and detail
Drill capability
Knowledge discovery in
databases
 KDD is the process of identifying valid,potentially
useful and understandable patterns & relationships
in data
 Knowledge = patterns & relationships
knowledge discovery =
data preparation + data mining + evaluation/interpretation of discovered
patterns/relationships
 Nowadays,  KDD = data mining

Knowledge Discovery in Database
Environment (Stages)
 There are six stages of KDD which are:
 Data selection
 Cleaning
 Enrichment
 Coding
 Data mining
 reporting

Data Mining: A KDD Process
Pattern Evaluation
 Data mining: the core of
knowledge discovery process.
Data Mining
Task-relevant Data
Data Selection
Warehouse
Data Cleaning
Data Integration

Databases
KDD : Data selection
 Data Selection
 It is the first stage of KDD process in which we collect and
select the data set or database required to work with
 Data sets are obtained from operational databases
 Obtaining information from centralized databases can be

difficult, reasons may be:
 Data set may need conversion from one format to another
 Eg: Excel files to access files

KDD: Data selection
 Different quality of data in different parts are available
 Making choice on right data is important
 Investigations should be made on any data warehouses
available in an organization.
 A well maintained DW helps to make data selection job
convenient by providing right data set necessary for
analysis.
 Data Cleaning
 This is the second stage of KDD.
 Data set obtained is never perfectly cleaned.
 We may not be aware of to what extent it is polluted.
KDD: Data selection
 Data in real world is dirty:
 Incomplete: lack attribute values
 Noisy: contains errors

 Human errors
 Not available when collected
 Not entered due to misunderstanding
 Malfunction of hardware/software
 Mistake data entry
 Inconsistent: contains discrepancies codes
 The cleaning phenomena should try to eliminate all the
above mentioned defects by the stage of de-duplication,
domain consistency, disambiguation
KDD …
 Enrichment
 Enrichment is the process of adding additional
information to the databases or accessing
additional databases to obtain extra information.
 Eg: an airline company might cooperate with

telephone company to enhance its marketing
policy. A telephone company maintains large
databases comprising the call behavior of
customers & create telephone profiles of the basis
of these data.
KDD …
 These telephone profiles could be used by airlines to
identify interesting new groups of target customers
 So data miners can collect all the necessary information

from additional bought – in databases.
 Obtaining information from other organizations may involve

some tedious procedures.
 Coding:
 Coding is one of the most important stage where further
cleaning and transformation of data is done.
KDD …
 Coding…
 It can range from simple SQL Queries to using
sophisticated high level languages depending upon
requirement.
 Some polluted records can be easily filtered out by using

SQL queries. (Eg: Records with most of the field empty can
be easily traced and removed)
 Coding is the creative activity which involves creative

transformation of data.
 It can be used to obtained more simpler form of the

complete, detailed database.
KDD: Coding
 Coding…
 Example: to the table of the magazine publisher,
we can apply following coding steps:
 Convert address to region (area codes)
 Birth date to age
 Divide income by 1000
 Divide credit by 1000
 Convert owners yes/no to 1/0
 Convert purchase date to month starting from 1990
 Perform filtering
KDD: Coding
 Coding: applying steps 1 to 6
Client Age income Credit Car H. Region Month Mag.

No owner Woner of purchas
purch ed
ase
203 20 18.5 17.8 0 0 1 52 Car
203 20 18.5 17.8 0 0 1 42 Music
209 25 36.0 26.6 1 0 1 Null Comic
203 20 18.5 17.8 0 0 1 48 house

KDD: Coding
 Coding: applying step 7
Client Age Incom Credit Car House Regio Car. House Sport Music Comic
NO e owner owner n Mag. Mag. Mag. Mag. Mag.
203 20 18.5 17.8 0 0 1 1 1 0 1 0
209 25 36.0 26.6 1 0 1 0 0 0 0 1

KDD: DataMining
 Data Mining:
 All the cleanings, transformations and enrichment are
performed on data, so that we can extract the most useful
information from it, and this is performed in data mining
stage of KDD.
 It consists of different rules, techniques, and algorithms
used for mining purpose.
 These are involved in performing following three tasks:
 Knowledge Engineering Tasks
 Classification tasks
 Problem solving tasks

Knowledge Engineering Tasks
KDD: DM
 DM…
Inductive Logic
Programming Different Algorithms
Concerned with
Different Tasks
*Association Rules
•K- nearest neighbor Genetic Algorithms
•Decision Trees
Classification Tasks Problem Solving Tasks

KDD: Data mining
 Knowledge engineering:
 is the process of finding right formal representation of
certain body of knowledge in order to represent it in a
knowledge based system
 Eg: Expert Systems (medical diagnostic system)
 Classification tasks:
 Classification is the process of dividing data into no. of
classes. Eg: class of customers
 Problem Solving Tasks:
 It involves finding solutions of remedies to the problems
that arise. Eg: why are people not going to cinema hall?
KDD…
 For finding useful patterns in databases, it is
necessary to choose right algorithms and right
tools.
 For choosing right data mining algorithms

following three points should be considered:
 Quality of input [No. of records, attributes,
numeric]
 Quality of output [yes/no results, statistics]
 Performance [CPU load]
KDD: Reporting
 This stage involves documenting the results
obtained from learning algorithms.
 Any report writer or graphical tools can be used
 It basically combines two functions:

 Analysis of results obtained from mining.
 Application of results to new data
 Different data visualization tools like scatter

diagrams available for showing different patterns or
clusters of data can be used.

Class15 - Data Warehousing

Uploaded by

Copyright:

Available Formats

Class15 - Data Warehousing

Uploaded by

Document Information

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

Class15 - Data Warehousing

Uploaded by

Copyright:

Available Formats

Data Warehousing

of data for decision makers, not on daily

CH#2, Data Warehousing By: Babu Ram Dawadi

CH#2, Data Warehousing By: Babu Ram Dawadi

CH#2, Data Warehousing

  Avoid constant growth

 Control response time (latency)

PERSONNEL STEP 2: Question the Data Warehouse

STEP 3: Do something DECISIONS

 By date, business type, geography, …

 By tapping into the operational database,

 Data analysis can provide information about

CH#2, Data Warehousing By: Babu Ram Dawadi

CH#2, Data Warehousing By: Babu Ram Dawadi

CH#2, Data Warehousing By: Babu Ram Dawadi

 Decision support system (DSS):

CH#2, Data Warehousing By: Babu Ram Dawadi

 Definition: Integrated, Subject-Oriented,

CH#2, Data Warehousing By: Babu Ram Dawadi

CH#2, Data Warehousing By: Babu Ram Dawadi

CH#2, Data Warehousing By: Babu Ram Dawadi

CH#2, Data Warehousing By: Babu Ram Dawadi

Example: Three dimensions By:

Mobiles Fax Standard

 A fact table stores measures as well as keys

 Dimensions are perspectives with respect to which

 A star schema defines a fact table and its

Fact table view: Multi-dimensional cube:

CH#2, Data Warehousing By: Babu Ram Dawadi

Han: Dataware Houses and OLAP 39

Drill Total Sales Drill

Drill Total Sales Drill

CH#2, Data Warehousing By: Babu Ram Dawadi

location by city location y country

measure is sales Drill down 2002

CH#2, Data Warehousing By: Babu Ram Dawadi

More detailed: 43go down in the granularity hierarchy

CH#2, Data Warehousing By: Babu Ram Dawadi

 The lattice of cuboids forms a data cube.

 The cuboid holding the lowest level of summarization is

 The top most 0-D cuboid, which holds the highest-level of

time item location supplier

time,item time,location item,location location,supplier

 Modeling data warehouses: dimensions & measures

branch location_key location to_location

 Relatively frequent updates

 Transactions access only a small fraction of the database

 On Line Analytic Processing -- OLAP

 Infrequent updates (Load)

 Transactions access a large fraction of the database

CH#2, Data Warehousing By: Babu Ram Dawadi

 Business intelligence involves the gathering, management, and

 Organizations can then make more strategic decisions about

CH#2, Data Warehousing By: Babu Ram Dawadi

Analysis Complexity & Value

Stage 1 Stage 2 Stage 3 Stage 4 Stage 5

Data to Information to Decisions

Data Information Access

On-line Updates Data Warehouse Query & Reporting

 Nowadays,  KDD = data mining