MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),

MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

1. OLTP - Online Transactional Processing System

OLTP stands for Online Transaction Processing, and it refers to a type of database system and
processing that focuses on managing and executing individual transactions in real-time. OLTP is
designed to handle high volumes of relatively simple, short, and rapid database transactions, such as
inserting, updating, or deleting records in a database. It serves as the backbone for the day-to-day
operations of various applications and systems, such as point-of-sale systems, airline reservations,
banking transactions, e-commerce websites, and more.
1.1 Queries that OLTP System can Process
Let us consider an example of supermarket store POS system, which is an OLTP system. Given below
are a set of queries that a typical OLTP system is capable of responding to:
• Search for a particular customer’s record.
• Retrieve the product description and unit price of a particular product.
• Filter all the products with a unit price equal to or above some price.
• Filter all the products supplied by a particular supplier.
• Search and display the record of a particular supplier.

1.2 Characteristics of OLTP System

Following are the few characteristics of a typical OLTP System:
(1) Real-time processing: OLTP systems are optimized for quick and immediate response to user
interactions. Transactions are processed in real-time, usually within milliseconds, to provide users
with timely information and ensure the data's accuracy and consistency.
(2) Concurrent users: OLTP systems are designed to support multiple concurrent users performing
simultaneous transactions. These users could be customers, employees, or other applications
interacting with the database.
(3) Short & simple transactions: OLTP transactions are typically short and involve simple read or
write operations on small amounts of data. Each transaction involves a specific business operation
like updating a customer's address or processing a payment.
(4) Data integrity: Maintaining data integrity is crucial in OLTP systems. These systems enforce data
validation and constraints to ensure that the database remains consistent and free from errors.
Page 1
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

(5) Normalized data structure: OLTP databases are often structured using normalization techniques
to minimize data redundancy and improve data integrity.
(6) High availability: OLTP systems require high availability to ensure continuous access to critical
data. Downtime can disrupt business operations and lead to data inconsistencies.
(7) ACID properties: OLTP systems follow the ACID (Atomicity, Consistency, Isolation,
Durability) properties to ensure that database transactions are executed reliably and safely.

1.3 Advantages of an OLTP System

(1) Simplicity: It is designed typically for use by clerks, cahiers, clients, etc.
(2) Fast Query Processing: OLTP databases are designed for quick query processing, making them
suitable for operations that involve retrieving specific records or updating individual data entries.
(3) Efficiency: It allows its users to read, write and delete data quickly.
(4) Support for Business Processes: OLTP systems are the backbone of various business processes,
supporting order processing, inventory management, customer management, and other day-to-day
operational tasks.
(5) Scalability: OLTP systems can be designed and scaled to handle increasing transaction volumes
as businesses grow. This scalability ensures that the system can accommodate future demands
without sacrificing performance.

1.4 Challenges of OLTP systems

(1) Performance: OLTP systems must handle a large number of concurrent transactions, which can
pose performance challenges. Optimizing the system for quick response times is essential to
prevent bottlenecks.
(2) Scalability: As businesses grow and the volume of transactions increases, OLTP systems need to
scale efficiently to accommodate the increased load.
(3) Data integrity and consistency: Maintaining data integrity in a rapidly changing environment
with concurrent transactions can be challenging. Ensuring that data remains consistent and free
from conflicts is critical.
(4) Reporting and analytics: OLTP databases are optimized for transaction processing and may not
be the most suitable choice for complex reporting and analytics tasks.
Page 2
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

To handle complex reporting and analytical requirements, organizations often use separate systems known
as Online Analytical Processing (OLAP) systems or data warehouses. These systems are optimized for
data analysis and decision-making, complementing the real-time transactional capabilities of OLTP
databases.
2. OLAP – Online Analytical Processing System

OLAP (Online Analytical Processing) is a technology used for multidimensional data analysis, enabling
users to perform complex and interactive data exploration and gain insights from large volumes of data.
Unlike OLTP (Online Transaction Processing), which focuses on handling individual transactions in real-
time, OLAP is designed to support analytical queries that involve aggregations, drill-downs, and slicing
and dicing of data to gain a deeper understanding of business trends and performance.

Key characteristics of OLAP include:

OLAP (Online Analytical Processing) systems possess several key characteristics that distinguish them
from OLTP (Online Transaction Processing) systems and make them ideal for analytical tasks and data
exploration. Here are the main characteristics of OLAP:

Multidimensional Data Model: OLAP systems use a multidimensional data model, represented as data
cubes. Data cubes organize information into dimensions and measures. Dimensions represent the different
attributes or perspectives of the data (e.g., time, geography, product categories), while measures are the
numeric values that are being analyzed (e.g., sales revenue, quantity sold).

Dimension Hierarchies: Within each dimension, OLAP systems define hierarchical relationships,
allowing data to be structured at different levels of granularity. For example, a time dimension might have
hierarchies at the year, quarter, month, and day levels.

Page 3
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

Aggregation and Summarization: OLAP systems support data aggregation and summarization along
multiple dimensions. Users can quickly view data at higher-level summaries or drill down to more detailed
levels of information.

Fast Query Performance: OLAP systems are optimized for rapid query performance, enabling users to
interactively explore data and obtain results in real-time. Aggregated data and pre-calculated measures
contribute to quick response times.

Ad Hoc Analysis: OLAP systems allow users to perform ad hoc analysis, where they can create custom
queries, apply filters, and manipulate data on the fly to explore different scenarios and derive insights.

Slicing and Dicing: Slicing involves selecting a specific subset of data from a cube by fixing the value of
one or more dimensions. Dicing, on the other hand, involves viewing data from multiple dimensions
simultaneously, providing a more comprehensive view of the data.

Drill-Down and Roll-Up: OLAP systems support drill-down and roll-up operations. Drill-down involves
navigating from higher-level summaries to more detailed data, while roll-up allows users to view data at
higher-level aggregations.

Online Response: OLAP systems are designed to provide real-time, interactive responses to user queries.
Users can explore data and make decisions without significant delays.

Business Intelligence and Decision Support: OLAP is a core component of business intelligence (BI) and
decision support systems. It helps users analyze data, identify trends, and gain insights to make informed
business decisions.

Integration with Data Visualization: OLAP systems often integrate with data visualization tools, allowing
users to create interactive and visually appealing charts, graphs, and dashboards to present analytical
findings.
Page 4
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

Analytical Functions: OLAP systems provide various analytical functions, such as ranking, forecasting,
time series analysis, and advanced calculations, to support complex data analysis.

Support for Historical Data: OLAP systems can handle historical data, making it possible to analyze trends
and changes over time.
These characteristics collectively enable OLAP systems to efficiently process analytical queries, explore
data from different angles, and support strategic decision-making and planning in organizations.

Difference between OLTP and OLAP

Page 5
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

3. Types of OLAP Systems

OLAP (Online Analytical Processing) can be categorized into three main types based on the way data is
stored and processed. These types are:

3.1 Multidimensional OLAP (MOLAP):

MOLAP systems store data in multidimensional cubes, where data is pre-aggregated and organized along
multiple dimensions. The cubes are typically created during the data processing phase, and they provide
rapid query performance for analytical tasks.
Key features of MOLAP include:
(1) Fast Query Performance: MOLAP cubes allow for quick response times as data is pre-
aggregated, enabling users to explore data interactively.
(2) Efficient Storage: Data is stored in a highly compressed format, optimizing storage requirements.
(3) Multidimensional Data Model: MOLAP supports multidimensional data models, allowing users
to analyze data from different perspectives.

Examples of MOLAP systems include Microsoft Analysis Services (part of Microsoft SQL Server), IBM
Cognos TM1, and SAP BW (Business Warehouse).

Page 6
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

While Multidimensional OLAP (MOLAP) has several advantages, it also comes with some disadvantages
that organizations should consider when choosing an OLAP technology. Here are some of the main
disadvantages of MOLAP:
(1) Cube Size and Storage Requirements: MOLAP cubes can grow significantly in size, especially
when dealing with large datasets and numerous dimensions. This can lead to increased storage
requirements, potentially making it more expensive to maintain and manage the infrastructure.
(2) Cube Processing Time: Building and processing MOLAP cubes can be time-consuming,
particularly for complex datasets. Generating and updating the cubes may take a considerable
amount of time, resulting in delays before users can access the most recent data.
(3) Cube Refresh Frequency: The data in MOLAP cubes may not always be up-to-date in real-time.
To keep the cubes current, they need to be refreshed at regular intervals. The frequency of refreshes
depends on the organization's data update rate and business needs.
(4) Aggregation Limitations: MOLAP relies on pre-aggregated data for faster query response times.
However, creating all possible aggregations for large datasets may not be feasible due to cube size
and processing constraints. This can lead to performance issues for certain analytical queries that
require aggregations not available in the cube.
(5) Data Complexity: While MOLAP is well-suited for relatively simple and regular data structures,
it may struggle with handling highly complex data models that involve numerous hierarchies,
cross-dimensional calculations, or sparse data.
(6) Scalability: MOLAP systems may face scalability challenges when dealing with very large
datasets or when the number of dimensions and hierarchies increases significantly. Scaling up the
infrastructure to support growing data volumes and complexity may be more demanding.
(7) Storage Redundancy: In MOLAP, data is pre-aggregated and stored in multidimensional cubes.
This pre-aggregation can lead to data redundancy, as the same data may be stored in multiple
aggregation levels.
(8) Data Consistency and Synchronization: MOLAP cubes need to be synchronized with the source
data to ensure consistency. If there are discrepancies between the source data and the MOLAP
cubes, it can lead to incorrect or outdated analytical results.

Page 7
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

(9) Cube Design Complexity: Designing an efficient MOLAP cube can be complex, especially for
large and diverse datasets. Creating an optimal cube structure that supports a wide range of
analytical queries requires careful planning and optimization.

Organizations should carefully assess their specific analytical requirements, data volumes, update
frequency, and performance expectations before opting for MOLAP as their OLAP solution. In some
cases, alternative OLAP approaches like ROLAP or HOLAP might be more suitable to address specific
challenges or business needs.

3.2 Relational OLAP (ROLAP):

ROLAP systems store data in traditional relational databases, and they perform on-the-fly calculations
during query execution. Instead of pre-aggregating data, ROLAP systems leverage the power of the
underlying relational database management system (RDBMS) to handle analytical tasks.

Key features of ROLAP include:

(1) Flexibility: ROLAP systems can handle large datasets, and they can adapt to changing business
requirements.
(2) Support for Complex Queries: ROLAP can perform complex calculations using SQL and leverage
database optimization techniques.

Examples of ROLAP systems include Oracle OLAP, Microsoft SQL Server Analysis Services
(Tabular model), and SAS OLAP Server.
Advantages of ROLAP:
Relational OLAP (ROLAP) offers several advantages that make it a valuable choice for certain
analytical requirements and business scenarios. Here are the key advantages of ROLAP:

(1) Scalability: ROLAP systems can handle large volumes of data and support complex analytical queries
without the limitations imposed by cube size or storage requirements. This scalability makes ROLAP
well-suited for organizations with extensive and diverse datasets.

Page 8
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

(2) Real-Time Data Access: Unlike MOLAP, which requires cube processing and refreshing, ROLAP
allows users to access data directly from the underlying relational database in real-time. This means
that users can work with the most up-to-date data, making ROLAP ideal for applications that require
real-time analytics.
(3) Flexible Data Model: ROLAP can accommodate changes and modifications to the data model
without the need to rebuild the entire cube. This flexibility allows organizations to adapt to evolving
business requirements and data structures.
(4) Support for Complex Queries: ROLAP leverages the power of SQL and the capabilities of the
underlying relational database management system (RDBMS) to handle complex analytical queries.
This includes support for advanced calculations, joins, and aggregations.
(5) Centralized Data Storage: ROLAP systems use a centralized relational database as the primary data
store. This ensures data consistency and integrity across the organization since there is a single source
of truth for all analytical data.
(6) Integration with Existing Systems: ROLAP can leverage the existing relational database
infrastructure, making it easier to integrate with various data sources and other business applications.
It eliminates the need for a separate cube-building process.
(7) Reduced Data Redundancy: ROLAP avoids data redundancy as it directly queries the underlying
relational database. This reduces storage requirements and ensures that data changes are immediately
reflected in analytical results.
(8) Better Handling of Sparse Data: ROLAP is more efficient in handling sparse data, where certain
combinations of dimensions have no data. It avoids the need to store empty cubes or aggregations,
making it more storage-efficient for datasets with sparsity.
(9) Security and Access Control: ROLAP systems can leverage the security features of the underlying
relational database, ensuring proper access control to sensitive data. This provides a robust security
framework for analytical data.
(10) Integrated Data Sources: ROLAP can access data from various sources, including multiple
relational databases and even data stored in different formats, allowing for a more comprehensive and
integrated view of the data.

Page 9
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

ROLAP's flexibility, real-time data access, and ability to handle complex queries make it an excellent
choice for organizations that require on-the-fly data analysis, access to up-to-date information, and
the ability to adapt quickly to changing business needs. ROLAP is particularly well-suited for
applications where data volumes are large, and analytical requirements demand frequent updates and
interactive exploration of data.

3.3 Hybrid OLAP (HOLAP):

HOLAP systems combine elements of both MOLAP and ROLAP to leverage the advantages of each
approach. HOLAP solutions store some data in pre-aggregated multidimensional cubes (similar to
MOLAP) while keeping the rest of the data in the relational database (similar to ROLAP).

Key features of HOLAP include:

(1) Balance of Performance and Flexibility: HOLAP strikes a balance between fast query
performance and the ability to handle large datasets and complex queries.
(2) Separation of Data: The system can segregate frequently accessed data in multidimensional
cubes for faster processing, while less frequently accessed data remains in the relational database.

Examples of HOLAP systems include Microsoft SQL Server Analysis Services (Hybrid model) and
Oracle OLAP.

Page 10
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

The choice between MOLAP, ROLAP, or HOLAP depends on factors such as the size of the dataset,
query complexity, performance requirements, and the organization's specific business needs. Each type
of OLAP system has its strengths and can be suitable for different use cases within an organization.

Data Warehouse
A data warehouse is a centralized, integrated, and subject-oriented repository that stores large volumes of
historical and current data from various sources within an organization. It is designed to support business
intelligence (BI) and decision-making processes by providing a unified view of data from disparate
systems and transforming it into a format suitable for analytical purposes.

Bill Inmon, a prominent figure in the field of data warehousing, has defined four key properties or
characteristics of a data warehouse. These properties are often referred to as "Inmon's Building Blocks"
and are foundational principles for designing and implementing a data warehouse. The four properties
are:

Page 11
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

Subject-Oriented:
Inmon's first property emphasizes that a data warehouse should be organized and focused around business
subjects or topics that are relevant to the organization. Each subject represents a specific area of interest,
such as sales, customers, products, finance, or inventory. By structuring the data warehouse around these
subject areas, it becomes easier for users to access and analyze data based on their specific business
requirements.

Integrated:
The second property of a data warehouse is integration. Inmon stresses the importance of integrating data
from various sources and operational systems into a single, consistent view. This integration process
involves standardizing data formats, resolving discrepancies, and consolidating data from disparate
sources. The integrated data ensures that users can obtain a unified and accurate view of information,
enabling more reliable and meaningful analysis.

Time-Variant:
Inmon's third property highlights the temporal aspect of data warehousing. A data warehouse should store
historical data over time, allowing users to analyze and track changes and trends. Historical data is
essential for identifying patterns, making comparisons, and conducting trend analysis, all of which are
valuable for decision-making and strategic planning.

Non-Volatile:
The fourth property emphasizes the stability and permanence of data in a data warehouse. Once data is
loaded into the data warehouse, it becomes non-volatile, meaning that it is not modified or updated further.
This ensures data consistency and integrity, as historical records remain unchanged and are not subject to
modifications due to ongoing transactions or operational updates.

Inmon's building blocks provide a solid foundation for designing data warehouses that meet the analytical
needs of organizations. By following these principles, data warehouses can deliver a unified, reliable, and
Page 12
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

historical view of data, enabling business users to make informed decisions and gain valuable insights
from their data.

Role of Data Warehousing in Business Intelligence

Data warehousing plays a crucial role in Business Intelligence (BI) by serving as a central repository for
integrated, historical, and well-organized data. It acts as the foundation for BI systems, providing a
structured and efficient environment for data analysis, reporting, and decision-making. The role of data
warehousing in business intelligence includes the following key aspects:

Data Integration: Data warehousing brings together data from various sources, such as operational
systems, databases, applications, and external data feeds. It performs the critical task of data integration,
where data from disparate sources is transformed, cleaned, and standardized to ensure consistency and
accuracy.

Historical Data Storage: Data warehouses store historical data, preserving a record of past business
operations and activities. This historical data is essential for trend analysis, identifying patterns, and
making comparisons over time, enabling users to gain insights into long-term business performance.

Performance Optimization: Data warehousing enhances query performance by using optimized data
structures like star schemas or snowflake schemas. These structures are designed to facilitate rapid
querying and reporting, allowing users to obtain results quickly even from large datasets.

Support for Complex Queries: BI systems often require complex analytical queries involving
aggregations, drill-downs, and multi-dimensional analysis. Data warehousing is designed to efficiently
handle these types of queries, enabling users to explore data from multiple angles.

Page 13
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

Data Consistency and Data Quality: Data warehousing ensures data consistency by integrating and
consolidating data from various sources. It also plays a role in data quality management, as it provides a
platform for data validation, cleansing, and error correction.

Data Transformation and ETL: Data warehousing involves Extract, Transform, Load (ETL) processes,
where data is extracted from source systems, transformed to conform to the data warehouse schema, and
then loaded into the warehouse. This process ensures that data is optimized for analytical purposes.

Single Source of Truth: Data warehouses serve as a single source of truth for the organization, providing
a unified view of data across departments and applications. This consistency fosters data-driven decision-
making and avoids data discrepancies.

Business Performance Management: Data warehousing supports business performance management by

providing the necessary historical data and KPI metrics to monitor and evaluate business performance
against goals and objectives.

Data Security and Governance: Data warehousing incorporates data security measures to protect sensitive
information. It also supports data governance practices, ensuring compliance with regulations and data
usage policies.

Business Intelligence and Reporting: Data warehouses serve as the backend for BI tools and reporting
platforms. They enable BI applications to access and analyze data efficiently, creating meaningful reports,
dashboards, and data visualizations for business users.

Page 14
 MAHARISHI MARKANDESHWAR (DEEMED TO BE UNIVERSITY),
MULLANA (AMBALA)
Programme: MCA
Course: MCA-101: Business Intelligence & its applications

In summary, data warehousing is the backbone of Business Intelligence, facilitating efficient data
management, integration, and analysis. It ensures that businesses have access to reliable, timely, and
actionable information, empowering them to make informed decisions and gain a competitive advantage
in the market.`

Page 15