The Data Warehouse Chapter 6
6.1 Operational Databases = transactional database designed to process individual transaction quickly and efficiently On-Line Transactional Processing (OLTP) Data Warehouse
Building a database: Data Modeling Normalization One-to-One Relationships One-to-Many Relationships Many-to-Many Relationships ERD (Entity Relationship Diagram)
Figure 6.1 A simple entity- relationship diagram
Normalization First Normal Form (atomic value) Second Normal Form (No 부분종속 ) R (A, B, C, D, E) Third Normal Form (No 이전종속 ) R (A, B, C, D, E)
The Relational Model 주문서 ( 주문번호, 주문일, 고객번호, 고객명, 주소, 제품번호, 제품명, 수량, 단가 ) 주 문 서 주문번호 : 주문일 : 고객번호 : 고객명 : 주소 : 제품번호 제품명 수량 단가 금액 1111 MP3 2 60, , 공 CD 3 10,000 30,000 합계 : 150,000
6.2 Data Warehouse Design OLTP Data Warehouse Process Oriented Subject Oriented Normalized Denormalized Day-to-day operation Historical Constant Update Not subject to change (read only) Lowest level of granularity Design issue
Figure 6.2 A data warehouse process model
Structuring the Data Warehouse: Fact Table ( dimension key + fact ) Dimension Tables ( Not Normalized, Slowly Changing Dimensions ) (1)Multidimensional Database (2)Relational Database Multidimensional Format Star Schema
Figure 6.3 A star schema for credit cared purchases
The Multidimensionality of the Star Schema Figure 6.4 Dimensions of the fact table shown in Figure 6.3
Additional Relational Schemas Snowflake Schema Dimension tables are further subdivided Constellation Schema Sharing dimensions
Figure 6.5 A constellation schema for credit card purchases and promotions
Decision Support: Analyzing the Warehouse Data Reporting Data Analyzing Data (multidimensional data analysis tool) Knowledge Discovery (through data mining)
6.3 On-line Analytical Processing (OLAP) - Query based methodology - Supports data analysis in multidimensional environment - Storage methods (1) Relational data store Star Schema (2) Multidimensional array data store
OLAP Operations Slice – A single dimension operation Dice – A multidimensional operation Roll-up – A higher level of generalization Drill-down – A greater level of detail Rotation – View data from a new perspective
Figure 6.6 A multidemensional cube for credit card purchases
Concept Hierarchy A mapping that allows attributes to be viewed from varying levels of detail.
Figure 6.8 Rolling up from months to quarters
6.4 Excel Pivot Tables for Multidimensional Data Analysis
Figure 6.15 A credit card promotion cube
Figure 6.16 A pivot table with page variables for credit card promotions