UNIT 1 | Data Warehousing and Data Mining Notes | AKTU Notes

1. Overview of Data Warehousing

Data warehousing is a system used for collecting, storing, and managing large amounts of data from different sources so that organizations can analyze it and make better decisions.

A data warehouse is mainly used for business intelligence (BI), reporting, and analysis rather than daily transactions.

Main Purposes

• Integrate data from multiple sources
• Store historical data
• Support decision making
• Improve business analysis

Hinglish Explanation

Data warehousing ek central storage system hota hai jisme different sources ka data collect karke store kiya jata hai analysis ke liye.

2. Definition of Data Warehouse

“A Data Warehouse is a subject-oriented, integrated, time-variant and non-volatile collection of data used to support management decision making.”

Key Characteristics

Subject Oriented

Data specific subject par organize hota hai (sales, customer, product).

Integrated

Different sources ka data ek common format me convert kiya jata hai.

Time Variant

Historical data store hota hai.

Non-Volatile

Data mostly read-only hota hai aur rarely update hota hai.

3. Components of Data Warehouse

1. Data Sources

• Databases
• Files
• ERP systems
• CRM systems

2. ETL Process

• Extract – Source se data lena
• Transform – Data clean aur convert karna
• Load – Data warehouse me store karna

3. Data Warehouse Storage

Central repository jahan processed data store hota hai.

4. Metadata

Metadata ka matlab hai data about data (table structure, data source information).

5. Front End Tools

• Reporting tools
• Data mining tools
• OLAP tools

4. Building a Data Warehouse

1. Requirement Analysis
2. Data Source Identification
3. Data Cleaning
4. Data Transformation
5. Data Loading
6. Data Analysis Tools Integration

5. Warehouse Database

Warehouse database ek large database system hota hai jo specially analysis aur reporting ke liye design kiya jata hai.

Features

• Huge historical data
• Optimized for queries
• Supports OLAP operations

6. Mapping Data Warehouse to Multiprocessor Architecture

Large data warehouses ko handle karne ke liye multiprocessor systems use kiye jate hain jisme multiple CPUs data process karte hain.

Types

• SMP (Symmetric Multiprocessing)
• MPP (Massively Parallel Processing)

7. Difference Between Database System and Data Warehouse

Feature	Database System	Data Warehouse
Purpose	Transaction processing	Data analysis
Data	Current data	Historical data
Updates	Frequent updates	Rare updates
Users	Operational staff	Managers and analysts
Queries	Simple queries	Complex queries

8. Multidimensional Data Model

Multidimensional model data ko multiple dimensions me represent karta hai analysis ke liye.

Example Dimensions

• Time
• Location
• Product

Components

• Fact Table – Numerical values (sales amount)
• Dimension Table – Descriptive attributes (time, product)

9. Data Cubes

Data cube ek multidimensional structure hota hai jo data ko multiple dimensions me represent karta hai.

Example Dimensions

• Product
• Time
• Location

Operations

• Slice
• Dice
• Roll-up
• Drill-down

10. Schema Concepts

Star Schema

• One fact table
• Multiple dimension tables
• Simple design
• Fast query performance

Snowflake Schema

• Dimension tables normalized
• More complex structure
• Less redundancy

Fact Constellation Schema

• Multiple fact tables
• Shared dimension tables
• Also called Galaxy Schema

Conclusion

Data warehousing helps organizations store and analyze large amounts of data from multiple sources. It supports decision making through multidimensional models, data cubes, and schema designs like star, snowflake, and fact constellation.