- In Star schema , dimensional data (describing how data is commonly aggregated) is separated from Fact or event data (describing individual transactions in the business.)
- Fact Tables (Tall Table) contain basic transaction level data. E.g. Invoices in sales system. The commonly used factors for analysis of data exist as reference data in form of foreign key in these tables. e.g. Customer , Product , time , region in invoices.
- Dimension Tables contain data needed to place transaction along a particular transaction. E.g. Customer , Product , Region.
- In Star Schema , de-normalization is done combining all possible attributes of similar objects in single table and then basic independent dimensions or entities are identified and placed as dimension table.
- The benefits are improved query performance due to lesser joins and second is simplicity from end user point of view.