Business Intelligence and Data Analytics

13 important questions on Business Intelligence and Data Analytics

What is data warehousing?

Data warehousing is the design and exploitation of a data warehouse database with:
- data from operational (OLTP) databases
- and enriched with data from external sources

Organized around the most important subjects of a company instead of applications and systems

What are OLTP databases?

- designed to support daily operations
- ensure fast concurrent acces to data for: transactions, concurrency control, recovery techniques
- normalization crucial: must support heavy transaction loads, should prevent update anomalies

Therefore: poor performance for executing complex queries (joining many relational tables together)

What are the key properties of a data warehouse?

Time variant:
- data warehouse offers a series of historical snapshots

Non-volatile:
- new data are always added
- old data remain
  • Higher grades + faster learning
  • Never study anything twice
  • 100% sure, 100% understanding
Discover Study Smart

What are the differences betwen data warehouse database vs. OLTP database?

Data warehouse database
- support of tactical and strategic decisions
- integrated: internal and external data input
- non volatile: data only read, never changed or deleted
- time dependent: data can be analyzed over time
- subject oriented: data organized around business objects
- data are often aggregated

OLTP
- support of operational processes
- not integrated
- volatile
- time independent
- transaction oriented
- data are detailed

What is the multidimensional model?

- High level description of the data model of a data warehouse
- N dimensional space: data cube or hyper cube

Dimension level or granularity level
- determined by the level of detail for each dimension
- dimensions have a hierarchical nature

Facts = cells of the data cube
- measures (value in each cell)

What is the conceptual data warehouse design?

- for transactional databases: ER, UML
- no well established and universally adopted conceptual model for multidimensional data
- therefore, often data warehouse design is usually directly performed at the logical level

Note
- main goal of conceptual model: communication between stakeholders
- main goal of logical model: implementation

So for datawarehousing; often difficult exercise

What logical data warehousing designs are there?

Relational OLAP (ROLAP) *important*
- data stored in relational tables
- pre computed aggregations stored for performance
- large space required
- OLAP performed with complex SQL statements
- relational DBMS: well standardized + large storage

Multidimensional OLAP (MOLAP)
- data cubes stored in multidimensional arrays
- verry efficient OLAP operations: performant querying
- MOLAP enginers typically support less storage
- often used for smaller "data marts" (<10 dimensions)

Hybrid OLAP (HOLAP)
- combines both previous approaches
- large volumes of data stored in a relational database
- aggregations are kept in a seprate MOLAP store

What relational data warehouse designs are there?

Star schema
2, types of tables
- facts tables: normalized
- dimension tables: generally not normalized

Snowflake schema
- avoids redudancy by normalizing dimension tables

What is data staging?

Data staging area = operational data store
- database in which the extracted data undergoes successive modifications

Steps
- data cleaning
- data transformation
- load
- refresh

What is a data warehouse tier?
What architectural variations are there?

Dat a warehouse: business wide scope
Data mart: supporting the decisions of a specific group

1. Direct acces to the data warehouse by client tools
2. Virtual data warehouse = defines a set of views over operational databases, materialized for efficient acces

What types of data warehouse usages are there?

Verification-based tools
- ad hoc querying an reporting
- online analytival processing (OLAP0

Discovery-oriented tools (data mining)
- knowledge discovery: search for hidden patterns
- supports associations, constructing analytical models, performing classification and prediction, and presenting the mining results using visualization tools

What OLAP operations are there?

Roll-up
- from city -> country

Drill-down
- from year -> month

Slice
- from cube of cities to 1 city

Dice      
- from cube to 2 cities and 2 months

What advanced topics are there in data warehousing?

- Spatial data warehousing: Geopgraphic information systems (GISs)
- Trajectory data warehouses
- In memory database systems
- Real time data warehouses
- Semantic web technologies
- Multimedia data warehouses

The question on the page originate from the summary of the following study material:

  • A unique study and practice tool
  • Never study anything twice again
  • Get the grades you hope for
  • 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
Trustpilot Logo