Data Model Example

From requirement to STAR schema


Information Packages are used to determine the requirements.
Table is an exmple of an information package. Columns are the dimensions. Dimensions can be either categorical (like 'User' in Table 1) or hierarchical (like 'Time Period'). Important is to determine the grain-level. For instance, if we take the 'Time period' column we can choose a lower level of granularity ('Minutes') or a higher level of granularity ('Days').
The row 'Facts' are facts that are measured in the Sakai Warehouse.

Content Usage IP

Time period

Organization Unit

User

Content

Year

HE organization

First Name

content

Month

School

Last Name

 

Week

Study Programme

Student ID

 

Day

Site

 

 

Hour

 

 

 

Facts

Activity, Unique Visitor, Visits

Every column in the IP table generates a dimension table. Dimension tables tend to 'wide' and a relative small amount of records. The 'Fact' row generates the fact table. Fact tables tend to have a large amount of records.