CBIS 4120 Chapter 3+

Data staging

The process whereby data are organized and prepared for analysis

• Ensures data are consistent, relevant, free of ambiguities, and can be analyzed in a reasonable amount of time

Data modeling

The definition of the data entities and their relationships

• data is modeled using entities, attributes, relationships and constraints

Data for analysis are extracted from the sources and then temporarily stored in a

data staging area

• enables the data modeler to gather data from multiple sources before moving them to their final destination

Data targets

are repositories of data upon which analytics can be preformed

• Warehouses, data marts, data lakes, cubes, swamps

• Typically the data target is an informational (OLAP) system modeled as a star schema

OLAP systems often use

denormalized database structures

Denormalized database

one that was originally normalized to eliminate anomalies

• reduces amount of tables in a database

• In OLAP systems are ‘read only’ - cannot be modified

• often called data warehouse

Data warehouse

a database architecture that provides the persistent storage of harmonized, cleaned, and consolidated data

• relational

Can informational and transactional databases be combined into one data structure?

No, transactional systems are optimized for quick writing, whereas informational systems are read-optimized

• Serve different purposes

• Can lead to performance degradation

In-memory databases (IMDB)

• Enterprises can now meet the requirements of both their transactional and informational systems

• SAP HANA