Modern organisations run two very different kinds of data work. One kind keeps the daily business running — recording a bank deposit, issuing a rail ticket, updating a stock count. The other kind looks across years of that recorded history to answer questions like "which districts are our loans defaulting in?" or "what should we stock before Eid?". The first is served by ordinary transactional databases; the second is served by data warehouses, data mining and business intelligence (BI). This lesson ties the whole mixed-data chapter together and is a frequent source of one-mark MCQs and short-note questions in BCS, Bank (Officer/IT) and various IT-officer examinations.
Expect questions such as: OLTP vs OLAP, star vs snowflake schema, what does ETL stand for, which data-mining task is market-basket analysis, what is metadata, difference between full, incremental and differential backup, and horizontal vs vertical fragmentation. This lesson makes each of them answerable.
Learning objectives
- Distinguish a database, a data warehouse and a data mart, and know when each is used.
- Compare OLTP and OLAP systems across purpose, data, queries and design.
- Explain the data-warehouse architecture and the ETL process step by step.
- Draw and contrast star and snowflake schemas and define fact and dimension tables.
- Define metadata and its role in a warehouse.
- Describe the main data-mining tasks — association (market-basket), classification, clustering, prediction — and compute support, confidence and lift.
- Explain business intelligence: dashboards, KPIs and reporting, with a banking example.
- Contrast big data with a traditional data warehouse.
- Differentiate full, incremental and differential backup and relate them to recovery.
- Explain distributed databases and horizontal vs vertical fragmentation.