A
What is a Data Warehouse?
store historical business data
What is Extract?
first step in ETL where data is collected from different sources.
What is OLTP?
used by banks and ATMs for daily transactions.
What is Load?
I am the "L" in ETL.
What is MOLAP?
faster because my summaries are already calculated in cubes.
What is ROLAP?
handle huge datasets because I use relational databases.
What is HOLAP?
combine cubes with relational tables.
What is Data Mining?
discover hidden patterns from warehouse data.
What is a Star Schema?
a schema with one fact table connected to several dimension tables.
What is a Snowflake Schema?
the centralized fact table is connected to multiple dimensions. In the snowflake schema, dimensions are present in a normalized form in multiple related tables.
What is Roll-up?
Aggregating data to a higher level
What is Drill-down?
Moving down in the concept hierarchy
What is Slice?
In a slice method, one dimension is chosen, and a subcube is generated.
What is Dice?
A dice operation selects two or more dimensions, and subcubes are generated
What is Data Integration?
reports from multiple departments cannot be combined properly.
What is HOLAP?
Combines ROLAP + MOLAP.
What is Pivot
rotate the data cube to provide a different view.
What is MOLAP?
stored in multidimensional cubes for very fast queries.
What is ROLAP?
use SQL to answer analytical queries.
What is a Data Lake?
A data lake is a centralized repository designed to hold vast volumes of data in its native, raw format — be it structured, semi-structured, or unstructured.
What is Transformation?
convert different date formats into one standard format.
What is Data Cleansing?
improve data quality before analysis.
What is Time-Variant?
represent historical information.
What is a Multidimensional Data Model?
organize data into facts and dimensions.
Difference between ETL and ELT
ETL (Extract, Transform, Load)
1. Data is transformed before loading into the warehouse
2. Only cleaned and transformed data is stored
ELT (Extract, Load, Transform)
1. Data is transformed after loading into the warehouse
2. Raw data is stored first, then transformed when needed