Rapid Battle Cards 01–14
- BI vs Data Warehouse — BI = analysis + enabling technologies; DW = integrated historical analytical data environment/store.
- Data Warehouse vs Data Warehousing — environment/store vs processes that populate, govern and maintain it.
- EDW vs Data Mart — enterprise analytical integration vs subject/department/process subset.
- Inmon vs Kimball — central normalized enterprise integration layer vs dimensional process areas integrated through conformed dimensions/bus.
- Subject-oriented vs application-oriented — organized around durable business subjects vs one application/function.
- Integrated vs copied — reconciled keys/codes/definitions/representations vs source inconsistencies preserved unchanged.
- Time-variant vs current-valued — historical point-in-time states vs emphasis on current state.
- Non-volatile vs volatile — stable historical append/preservation vs frequent operational change.
- Atomic vs summarized — lowest useful analytical detail/flexibility vs aggregates optimized for particular questions/performance.
- Staging vs DW — intermediate preparation vs persistent integrated historical analytical storage.
- ODS vs DW — low-latency current/near-current/more volatile vs historical/stable.
- ODS vs Data Mart — integrated operational/current reporting vs targeted analytical subject/process presentation.
- Data Mart vs Cube — analytical subset/store vs OLAP multidimensional access structure.
- Fact vs Dimension — quantitative process measure vs descriptive analytical context.
Fast drill
- Five years of reproducible history → DW.
- Every-few-minutes integrated current status → ODS.
- Temporary cleanse/standardize before load → Staging.
- Revenue by Product by Month → Revenue fact; Product/Month dimensions.
Source: pp. 361–371.