CS5504: Business Intelligence Systems, Infrastructures and Technologies Assignment 2022

CS5504: Business Intelligence Systems, Infrastructures and Technologies
Assignment
SOLUTION

Introduction
BI (Business Intelligence) is a set of all the processes, architectures, and technologies used for
converting the raw data related to business into meaningful information actions.
Initially, a source of information is fed into a computer via a person. The database administrator
updates the files the customer gets feedback. Meanwhile, there is an online analytical processor
that stores all the files in a data-warehouse with the help of Business Intelligence. It can modify,
insert or inform a customer regarding the information.
Architectural Paradigms in a BI Context (Koutsoukos, CS5504/CS4504: Business)
Common Architectural Paradigm related issues to be tackled in a BI Infrastructure
implementation:
 Architecture (Logical and Physical) of each of: Data Warehouse/Data Marts/ODS/Staging
Area and Visualization Applications
 The architecture of the operational applications which are the data sources.
 ETL Tools Architecture
 Exploiting BI data Architecture of custom-built applications
Differences between BI and OLTP systems (Koutsoukos, CS5504/CS4504: Business)
Criteria Transactional Systems BI Systems (Data
Warehouse based)
Purpose Perform day-to-day
business operations
Support decision making
Usage OLTP systems support
predefined operations
Designed for ad hoc

(i.e. have a fixed, “a-
priori” known set of
operations in the data)
querying and analysis
Data Modifications s The operations that
end-users perform result
in direct data
modifications.
Data warehouse/BI data
is updated on a regular
basis using massive data
modification techniques
(ETL).
DB Schema Design Normalized schemas in
order to optimize the
efficiency/consistency of
“update/insert/delete”
data operations.
Combinations of
relational/normalized
and multi-dimensional
DB schemas (e.g. a star
schema) that optimize
(ad-hoc) querying and
data analysis.
Historical Data DB queries access only
some records
DB queries usually scan
thousands or millions of
rows
Size GBs TBs
In an OLAP application, the data is modeled as a (Star and SnowFlake Schema in Data
Warehousing, n.d.):
 Dimensional Modeling - Star Schema

 Dimensional Modeling - Snowflake schema
 Dimensional Modeling - Cube (Olap Cube)
A dimensional schema is a denormalized schema that follows the business model. Dimensional
schemas contain:
 dimension tables, for storing the attributes,
 fact tables, foreign keys of each of the dimension tables.
Following answers below justifies the above statements made.
Answer to question 1