Business Intelligence

Business Intelligence

• Set of concepts & methodologies to improve decision-making

• Can change raw data into significant & useful data for business purposes

• Goal is to improve decision making

Hans Peter Luhn

• Published “A Business Intelligence System” in 1958

• Father of Business Intelligence

Main Components:

• Data Source

• Data Mart / Data Warehouse

• Data Exploration

• Data Mining

• Optimization

• Decisions

Data Source

• Gather & consolidate from an array of primary & secondary sources

Data Mart / Data Warehouse

• Utilized ETL (Extract, Transform, Load)

• Centralized location

Data Exploration

• Offers essential resource for conducting passive analysis

• Techniques are referred to as passive

Data Mining

• Extracting valuable information & knowledge from data

• Aka Active Business Intelligence methodologies

Optimization

Improves decision-making by finding the most efficient & effective solution/s to a business problem

TYPES OF USERS OF BI:

1. Analysts (Data Analysts / Business Analysts)

2. Head / manager of the company

3. IT Engineer

4. Small Business Owner

5. Government Officials

Data Warehouse (definition)

• Relational database housed in a computer system in an on-premise data-center / cloud

DATA WAREHOUSE ARCHITECTURE:

• Database server

• Analytics engine

• BI & Reporting tools

Extract, Transform, Load (ETL)

• Set of process from gathering data up to storing of data in a DW

• Most appropriate for small data sets which require complex transformations

• Preferred when larger data sets & timelines is important

ETL pipeline

Traditional type of data pipeline for cleaning, enriching, and transforming data from a variety of source

Batch Processing

Cloud-Native

Open Source

Real-Time

PRIMARY TYPES OF ETL TOOLS:

Data Visualization Principles:

Define a Clear Purpose

Know the Audience

Use visual features to show the data properly

Line Charts

Values over Time

Bar Charts

Compare QUANTITATIVE DATA from several categories

Changes over time too, only when changes are significant

Scatter Plots

Values for 2 var. for a set of data; excellent for exploring relationships between the 2 sets

Pie Charts

Show parts of a whole; can’t display things like change overtime