UNIT - 3

INFORMATION ACCESS AND DELIVERY

UNIT - 3
INFORMATION ACCESS AND DELIVERY
 Matching information to the classes of users – Chapter 14

 OLAP in Data Warehouse – Chapter 15

 Data Warehouse deployment – Chapter 19

Book: Data Warehousing Fundamentals: A Comprehensive Guide for IT

2
Professionals. Paulraj Ponniah
INTRODUCTION
 Sources have been successfully identified.
 Extracted and transformed the source data.

 Best data design for the data warehouse

repository.
 Loaded the transformed and cleansed data into
your data warehouse database.

 NOW WHAT?
Best possible mechanism for information delivery
to your users – user interface for information
3
INTRODUCTION

 Who users are,

 What information they need,

 When and where they need the information, and

 In exactly what form they need the information

General classes of users of a typical warehouse and

the methods for providing information to them

4
INFORMATION DELIVERY
DATA WAREHOUSE VERSUS OPERATIONAL SYSTEMS

5
INFORMATION USAGE MODES
 Verification Mode:
 In this mode, the business user proposes a hypothesis
and asks a series of questions to either confirm or
repudiate it.
 E.g.- The marketing department goes to the data
warehouse with the hypothesis that the sales in the
South Region have increased. Information from the data
warehouse will help confirm the hypothesis.
 Discovery Mode:
 In this mode, the business analyst desires to discover
new patterns of customer behavior or product demands.
 The user does not have any preconceived notions of what
the result sets will indicate. 6
APPROACHES FOR INTERACTION
 Informational Approach:
 Users retrieve historical or current data with query and
reporting tools, and perform some standard statistical analysis.
The result sets may take the form of reports and charts.
 Analytical Approach:
 The users make use of the data warehouse for performing
analysis.
 They do the analysis along business dimensions using historical
summaries or detailed data.
 Data Mining Approach:
 Both the informational and analytical approaches work in the
verification mode.
 The data mining approach, however, works in the knowledge
7
discovery mode.
CLASSES OF USERS

8
INFORMATION DELIVERY FRAMEWORK

9
OLAP IN DATA WAREHOUSE
 DEMAND FOR OLAP
 Data warehouse is meant for substantial analysis using
the available data. This analysis leads to strategic
decisions.
 Multidimensional analysis –
 The users must have easy methods of performing complex
analysis along with several business dimensions.
 They need an environment that presents a multidimensional
view of data.
 Fast Access and powerful calculations
 Size of the query or the amount of time to receive the result
sets must be consistent irrespective of the query types.

How much revenue did the new Product X generate during the last three months, broken
down by individual months, in the South Central territory, by individual stores, broken
down by promotions, compared to estimates, and compared to the previous version of the11
product?
MULTIDIMENSIONAL DATA:
A UNIVERSITY SAMPLE DATA CUBE
Module Average Mark
Computing of Abraham in Year 1.
Business Avg
Abraham
Bridget Year 1
Caroline
Avg
Year 2

Time
Year 3

Avg

All, All, All

12

Students’ marks as a function of student, department, and year

EXAMPLE OF AN ANALYSIS SESSION

13
LIMITATIONS OF OTHER ANALYTICAL
METHODS

Report Writers Ability to generate reports Do not support multi

based on SQL calls dimensionality.
Spreadsheets Performs What if analysis Limitations on
multidimensional analysis
and complex calculations.
SQL Small retrievals Analysis for market
exploration and financial
forecasting.

14
 OLAP IS THE ANSWER

• OLAP enables analysts, executives and managers to

gain useful insights from data presentations.
• Can recognize metrics along several dimensions and
allow data to be viewed from different perspectives.
• Supports multidimensional analysis.
• Is able to drill down or roll up within each dimensions.
• Complements the use of other information delivery
techniques such as data mining.
• Provides fast responses.
• Designed for highly interactive analysis.
• Can be implemented on the web.
15
EXAMPLE: OLAP SESSION

16
OLAP DEFINITION

OLAP is a category of Software technology that

enables, managers, analysts and executives to gain
insight into data through fast, consistent,
interactive access in a wide variety of possible views
of information that has been transformed from raw
data to reflect the real dimensionality of the
enterprise as understood by the user.

17
OLAP CHARACTERISTICS

 Lets users have a multidimensional and logical

view of the data in a DW.
 Facilitates interactive querying and complex
analysis.
 Allows users to roll-up or drill down for greater
details.
 Perform intricate calculations.

 Results displayed in charts and graphs.

18
MAJOR FUNCTIONS AND FEATURES

19
DIMENSIONAL ANALYSIS - EXAMPLE

20
EXAMPLE: A THREE DIMENSIONAL DISPLAY

Pages: STORE dimension Columns: PRODUCT dimension

Hats Coats Jackets

Jan 200 480 350
Feb 210 480 390
Rows: TIME dimension

Mar 190 550 380

Apr 190 345 350
May 160 234 320
June 230 345 310
Jul 140 567 270
Aug 222 570 250
Sep 454 470 240
Oct 180 480 260
Nov 170 520 280 21
Dec 200 560 320
SAMPLE QUERIES

 Display the total sales of all products for past five years
in all stores
 Rows:-Year numbers 2000,1999,1998,99,97
 Columns:-Total sales for all products
 Page:-One store per page.

 Compare total sales for all stores, product by product,

between years 2000 and 1999
 Rows:-Year numbers 2000,1999
 Columns:-Total sales for all products
 Page:-All stores

22
MULTIDIMENSIONAL DOMAIN STRUCTURE(MDS)

• MDS is well suited to represent four dimensions.

• The data may be displayed on a spread sheet showing
metrics as columns, time as rows and product as pages.
This representation of a multidimension is referred to as
multidimensional domain structure(MDS)
• A hypercube is a general metaphor for representing
multidimensional data.
• The problem of accommodating four dimensions with
only 3 display groups can be resolved by combining
multiple logical dimensions within the same display
group. Metrics can be combined with products to display
as columns.
23
MDS – EXAMPLE FOR 3 DIMENSIONS

24
MDS – EXAMPLE FOR 4 DIMENSIONS

25
MDS – EXAMPLE FOR 4 DIMENSIONS

26
DRILL DOWN AND ROLL UP

27
 DRILL DOWN
The drill-down operation (also called roll-down) is the reverse operation
of roll-up. Drill-down is like zooming-in on the data cube. It navigates from
less detailed record to more detailed data.

28
29
ROLL-UP
The roll-up operation (also known as drill-up or aggregation
operation) performs aggregation on a data cube, by climbing down
concept hierarchies, i.e., dimension reduction.

30
31
 SLICING
The Slice operations perform a selection on one dimension of the given cube,
thus resulting in a subcube.

Food Outdoor CATEGORY_tot

Line Line al
Asia 59,728 151,174 210,902

Europe 97,580.5 213,304 310,884.5

North America 144,421.5 326,273 470,694.5

REGION_total 301,730 690,751 992,481

Slice

Food Line Outdoor Line CATEGORY_total

32
North America 144,421.5 326,273 470,694.5
33
DICING
The dice operation describes a subcube by operating a selection on two or more
dimension.

Food Line Outdoor Line CATEGORY_total

Canada 29,116.5 69,310 98,426.5
Mexico 12,743.5 24,284 37,027.5
United States 102,561.5 232,679 335,240.5

Dice

Food Line Outdoor Line

Mexico 12,743.5 24,284 34
United States 102,561.5 232,679
35
PIVOT

 The pivot operation is also called a rotation. Pivot is a

visualization operations which rotates the data axes in view to
provide an alternative presentation of the data. It may contain
swapping the rows and columns or moving one of the row-
dimensions into the column dimensions.

36
37
OLAP: USES AND BENEFITS

 Increased productivity of business managers,

executives and analysts
 Users can run their own analysis without IT
assistance.
 Faster delivery of applications.

 Ability to model real-world challenges with

business metrics and dimensions.

38
OLAP MODELS

 ROLAP – Relational online analytical processing.

 MOLAP – Multidimensional online analytical
processing.

39
ROLAP MODEL

 Data is stored as rows and columns in relational

form.
 Presents data to the users in the form of business
dimensions.
 Metadata layer supports the mapping of
dimensions to the relational tables.
 Unlike MOLAP model, static multidimensional
structures are not created and stored.
 Three distinct characteristics
 Supports all the basic OLAP features and functions
 Stores data in a relational form
 Supports some form of aggregation 40
ROLAP MODEL

41
THE MOLAP MODEL

 Data for analysis is stored in specialized multidimensional

databases.
 Pre-calculated and prefabricated multidimensional data cubes
are stored in multidimensional databases.
 The MOLAP engine in the application layer pushes a
multidimensional view of data from the MDDBs to the users.
 MDBMS are proprietary software systems.
 These systems provide the capability to consolidate and
fabricate summarized cubes during the process that loads
data into the MDDBs from the main data warehouse.
 The users who need summarized data enjoy fast response
times from the pre-consolidated data.
42
MOLAP MODEL

43
DIFFERENCE BETWEEN MOLAP AND ROLAP

MOLAP ROLAP

 Multidimensional  Relational OLAP.

OLAP.
 OLAP engine resides  OLAP engine resides
on a special server. in the desktop.
 MDDBs store data  The relational data is
in the form of presented as virtual
Multidimensional multidimensional data
hypercubes cubes.

44
ROLAP VERSES MOLAP
(DATA STORAGE)

ROLAP MOLAP
Stored as relational tables Stored as relational tables
Detailed and light summary Various summary data kept in proprietary
data available databases (MDDBs)
Very large data volumes Moderate data volumes
All data access from the Summary data access from MDDB
warehouse storage. detailed data access from warehouse.

45
ROLAP VERSES MOLAP
(UNDERLYING TECHNOLOGIES)

ROLAP MOLAP
Complex SQL is used to fetch Pre-fabricated data cubes by MOLAP
data from DW. engine. Proprietary technology to store
multidimensional views in arrays and not
tables.
ROLAP engine in analytical High speed matrix data retrieval.
server creates data cubes in a fly.
Multidimensional views by Spare matrix technology to manage data
presentation layer. sparsity in summaries.

46
ROLAP VERSES MOLAP
(FUNCTIONS AND FEATURES)

ROLAP MOLAP

Known environment and availability Large library of functions for complex

of many tools. calculations.
Limitations on complex analysis Easy analysis irrespective of the
functions. number of dimensions.
Drill-through to lowest level easier. Drill down and slice and dice
Drill-across is not always easy. capabilities.

47
DATA WAREHOUSE DEPLOYMENT

 Is the phase after construction.

 Main concerns in the phase relate to the users getting
the training, support, and the hardware and tools they
need to get into the data warehouse.
 There are 6 major activities in this phase.
⚫ Complete user acceptance
⚫ Perform initial loads
⚫ Get user desktops ready
⚫ Complete initial user training
⚫ Institute initial user support
⚫ Deploy in stages

48
PHASES OF DATA WAREHOUSE DEPLOYMENT

49
DATA WAREHOUSE DEPLOYMENT
(CONTD.)

 Complete user acceptance

⚫ Is an absolute necessary.
⚫ Testing of some predefined queries and reports also
done.
⚫ System performance also checked.
⚫ When successful, is a signal that the project is ready for
full deployment.
 Perform initial loads
⚫ In some cases it may run for many days.
 Get users desktops ready
⚫ What the users see and experience at their desktops is
what counts for them.
⚫ Special attention to be given to the installation of data
access tools, the n/w connections and the configuration of
the middle tiers. 50
DATA WAREHOUSE DEPLOYMENT
(CONTD.)
 Complete initial user training.
⚫ Users to be trained in the areas of basic database, data
storage concepts, features of data warehouse, contents of
DW, browsing through DW etc..
 Institute initial user support

51
DATA WAREHOUSE DEPLOYMENT
(CONTD.)

 Deploy in stages

52
Thank You

53