ONLINE ANALYTICAL

PROCESSING
 INTRODUCTION
The term OLAP was coined by E.F.Codd in 1993 , to refer to
a type of application that allows user to interactively analyze
data.
 It describes a class of applications that require multi-
dimensional analysis of business data.
OLAP systems enable managers and analysts to rapidly and
easily examine key performance data and perform powerful
comparison and trend analyses, even on very large data
volumes.
They can be used in a wide variety of business areas,
including sales and marketing analysis, financial reporting,
quality tracking, profitability analysis, manpower and
pricing applications and many others.
OLAP is a method of analysing data in multi-dimensional
format, often across multiple time periods, with the aim of
uncovering the business information concealed within the
data.
OLAP enables business users to gain an insight into the
business through interactive analysis of different views of
the business data that have been built up from the
operational systems.
OLAP is not a data warehousing methodology, it is an
integral part of a data warehousing solution.
OLAP comes in different shades depending upon the
underlying database structure and the location of the
majority of the analytical processing.
OLAP provides the facility to analyze the data held within
the data warehouse in a flexible manner.
 DEFINING OLAP

OLAP can be defined as the process of converting raw data into

business information through multi-dimensional analysis.
This enables analysts to identify business strengths and weaknesses,
business trends and the underlying causes of these trends.
It provides an insight into the business through the interactive
analysis of different views of business information that have been
built up from raw operating data and which reflects the business
users understanding of the business.
It allows the business users to understand the current business
position,and the factors contributing to that position, through
detailed analysis of the underlying information.
It also helps business users to identify important business trends
and opportunities through the analysis of historical data and future
projections in various what-if scenarios.
The OLAP application contains logic which includes:
Multi-dimensional data selection
Sub-setting of data.
Retrieval of data via the metadata layer.
Calculation formulas.
 KEY FEATURES OF OLAP

MULTI-DIMENSIONAL VIEWS
Managers typically look at financial data by scenario (actual vs budget), organisation,
line items and time; and at sales data by geography, channel and time.
A Multi-dimensional view of data provide more than the ability to “slice and dice”, it
provides the foundation for analytical processing through flexible access to
information.
Database design should not prejudice which operations can be performed on a
dimension or how rapidly those operations are performed.
Managers must be able to analyze data across any dimension, at any level of
aggregation, with equal functionality and ease.
OLAP software should support these views of data in a natural and responsive
fashion.
Whether a request is for the weekly sales of a product across all geographical areas
or the year-to-date sales in a city across all products, an OLAP system must have
consistent response times.
Managers should not be penalised for the complexity of their queries in either the
effort required to form a query or the amount of time required in receiving an
answer.
 COMPLEX CALCULATIONS
Key performance indicators often require involved algebraic equations.
Sales forecasting uses trend algorithms such as moving averages and percentage
growth.
Analyzing the sales and promotions of a given company and its competitors
requires modeling complex relationships among the players.
The real world is complex-the ability to model complex relationships is the key in
analytical processing applications.
OLAP software must provide a rich tool kit of powerful yet succinct computational
methods.
To make developers more efficient and business users more self sufficient, the
vehicle for implementing computational methods should be clear and non-
procedural.
If the method for creating the desired calculations is not clear, development time
and /or usage will suffer.
If the calculation method is procedural, changes to the system cannot be done in a
timely manner, which affects the just-in-time information.
The analytical processing systems are judged on their ability to create information
from data.
Example:- Calculation of Margin, Allocations and Trend Analysis.
 TIME INTELLIGENCE

Time is an integral component of almost any analytical

application.
True OLAP systems understand the sequential nature of
the time.
Business performance is always almost judged over time.
Time hierarchies is not always used in the same manner as
the other hierarchies.
Concept such as year-to-date and period over period
comparisons must be easily defined in an OLAP system.
OLAP systems must understand the concept of balance
over time.
 DIFFERENT STYLES OF OLAP

Increasingly, companies are implementing OLAP

architectures to support their analytical requirements
within data warehouse implementations.
Older OLAP applications have generally been
implemented as stand-alone applications that do not
allow easy sharing of data with other OLAP applications.
The newer OLAP tools are moving away from multi-
dimensional database management technology and
increasingly use relational database management
technology.
 The four major alternatives for implementing OLAP applications
are:-

Multi-Dimensional OLAP
Hybrid OLAP
Desktop OLAP
Relational OLAP
 MULTI-DIMENSIONAL OLAP

Based on multi-dimensional database architecture.

This stores data in a three dimensional data cube that is already in the
OLAP multi-dimensional format for “slicing and dicing” into analysis
view.
Multi-dimensional architecture provides performance benefits where
the data retrieval paths follow predefined structure of the data cubes.
To maximise performance and provide predictable response times for
complex analysis queries, the database is designed to support the
intended data analysis.
MOLAP is suitable for applications requiring only pre-defined analysis
on multiple dimensions.
Eg:- Sales by customer type by geographic area over time.
Not suitable for adhoc analysis.
 HYBRID OLAP

They primarily integrate specialized multi-dimensional data

storage with relational database management technology.
This allows businesses to link multi-dimensional data to the
underlying source data in a relational database.
When the analyst reaches the bottom of the multi-dimensional
hierarchy during the “drill-down” analysis, the application can
obtain the underlying detail data from the relational database.
At first glance, hybrid OLAP products seem to combine the better
points of both multi-dimensional and relational databases.
HOLAP is best suited to applications that require heavy analysis,
must provide predictable response times to resource intensive
queries, will have a small number of concurrent users (fewer than
20).
Not suitable for ad-hoc analysis.
 DESKTOP OLAP
It allows the user to perform limited analysis, directly against data held within
a relational database, while avoiding many of the problems that affect the
hybrid and relational OLAP styles.
These products provide multi-dimensional analysis features such as slice and
dice analysis capabilities.
These capabilities are achieved by creating a query to select data from the
operational relational database, importing the requested data onto the desktop
an placing it in a data cube (multi-dimensional array) and performing multi-
dimensional analysis against the data cube.
The data cubes can be stored and maintained on the desktop, thus eliminating
the overhead of building the data cube each time the data is analyzed.
The desktop OLAP is suitable for an enterprise that wants to provide pre-
defined analysis capabilities to business users without incurring the higher
purchase and maintenance cost of more functional products.
Not suitable where there is a requirement for flexible analysis or where
multiple users require access to a common view of the business data.
 RELATIONAL OLAP

Fastest growing area of OLAP technology, with new vendors entering the
market at an accelerating pace.
They are designed to operate directly on a data warehouse built on a
relational databases, through a comprehensive metadata layer, which
eliminates any need to create static multi-dimensional data cubes.
Multiple multi-dimensional views of the two dimensional relational data
can be created, as new requirements evolve without the need to restructure
the underlying database.
New entrants into this marketplace can provide this functionality easily , as
they are not constrained by existing product architectures.
ROLAP is suitable for situations where users require unrestricted analysis
of a large volume of data, different business areas require different multi-
dimensional views over the same data source, where there is a requirement
to drill down to a low level of detail without impacting on the operational
system etc
Not suitable where data storage is a limiting factor due to the data
redundancy.
 HOW TO CHOSE AN OLAP STYLE??

If only pre-defined multi-dimensional analysis is required, a multi-dimensional

OLAP approach is most suitable due to the peed of processing. But it will be
difficult to incorporate any great degree of ad-hoc analysis at a later stage.
If the requirement is for intensive analysis of large volumes of data by a small
number of concurrent users (fewer than 20) , then a hybrid OLAP solution may be
appropriate. Cost of this solution and the skill requirement are the main
disadvantage.
If the requirement is for limited pre-defined data analysis ,a desktop OLAP
application may provide a cost-effective solution. But it will be difficult to enhance
this to provide flexibility analysis at a later stage.
If there is a requirement for flexible querying, for eg where the business is
changing rapidly as in the electricity and gas markets currently, the inflexibility of
a multi-dimensional cube will be unsuitable. I this case ROLAP is the only option.
If the business requires unrestricted analysis of a large volume of data where there
is a requirement to drill down to a low level of detail , a ROLAP solution is most
appropriate.
If there is a requirement to move the analysis into an intranet, or the internet at a
later stage the heavy client-processing requirement of desktop OLAP is unsuitable.
 BENEFITS OF OLAP
OLAP applications increase the productivity of business managers,
developers and whole organizations.
OLAP enables managers to model problems that would be impossible
using less flexible systems with lengthy and inconsistent response times.
IT users also benefit from using the right OLAP software.
By using the software specifically designed for OLAP, developers can
deliver applications to business users faster and providing better service.
OLAP reduces the applications backlog still further by making the
business users self-sufficient enough to build their own models.
By using software designed for OLAP,IT reduces the query drag and
network traffic on transaction systems or the data warehouse.
By providing the ability to model real business problems and a more
efficient use of people, resources, OLAP enable the organization as a
whole to respond more quickly to market demand. Market
responsiveness in turn often yields improved revenue and profitability
 USES OF OLAP
Finance departments uses OLAP for applications such as budgetting, Activity based
costing, financial performance analysis and financial modelling.
Sales dept uses it for Sales analysis and forecasting.
Marketing dept uses it for market research analysis, sales forecasting , promotions
analysis, customer analysis and Market /Customer segmentation.
OLAP is also used for Production planning and Defect analysis.
OLAP has the ability to provide managers with the information they need to make
effective decisions about an organization’s strategic directions.
A truly flexible data model ensures that OLAP systems can respond to changing
business requirements as needed for effective decision making.
OLAP enable analysts, managers and executives to gain insight into data through
fast, consistent, interactive access to a wide variety of possible views of information.
OLAP transforms raw data so that it reflects the real dimensionality of the enterprise
as understood by the user.
While OLAP systems have the ability to answer the “who” and “what” questions , it is
their ability to answer the “what-if” and “why” that sets them apart from data
warehouses.
OLAP plays the role of a mediator to the various types of data sources and front end
interfaces.
MEDIATING ROLE OF OLAP SERVER

GRAHICAL STATISTICAL
SPREADSHEET
INTERFACE PACKAGE

OLAP SERVER

HIERARCHICAL
RDBMS FLAT FILES
DATABASE