Data-driven DSS Chapter 7
Types of Data-driven DSS Data warehouses Executive Information Systems Spatial DSS Online Analytical Processing
Source of data Internal data External data
Detail and Summary INV_NUMBERLINE_NUMBERP_CODELINE_UNITSLINE_PRICE Q2/P21$ HB1$ T2$ /QPD1$ QQ21$ Q2/P25$ T3$ HB2$ PVC23DRT12$ SM $ /QTY1$ HB1$ WRE-Q1$ Q2/P22$ T1$ PVC23DRT5$ WR3/TT33$ HB1$9.95 P_CODEUnitsSoldAvgPrice 13-Q2/P28$ QQ21$ /QTY1$ /QPD1$ HB5$ T6$ WRE-Q1$ PVC23DRT17$5.87 SM $6.99 WR3/TT33$ Drill-down
Dimensions P_DESCRIPTProductsSold AvgPri ce CUS_AREACO DE VendorStat e 1.25-in. metal screw, 253$ TN 7.25-in. pwr. saw blade6$ KY 7.25-in. pwr. saw blade2$ KY B&D cordless drill, 1/2-in.1$ FL B&D jigsaw, 12-in. blade1$ TN Claw hammer2$ TN Claw hammer3$ TN Hicut chain saw, 16 in.1$ TN Hrd. cloth, 1/4-in., 2x501$ GA Rat-tail file, 1/8-in. fine6$ KY Steel matting, 4'x8'x1/6",.5" mesh 3$ FL Dimension Dimensions
Slice and Dice P_DESCRIPT Products Sold AvgPriceCUS_AREACODE 1.25-in. metal screw, 253$ in. pwr. saw blade6$ in. pwr. saw blade2$ B&D cordless drill, 1/2- in. 1$ B&D jigsaw, 12-in. blade 1$ Claw hammer2$ Claw hammer3$ Hicut chain saw, 16 in.1$ Hrd. cloth, 1/4-in., 2x501$ Rat-tail file, 1/8-in. fine6$ Steel matting, 4'x8'x1/6",.5" mesh 3$ Viewing Data by Customer Area code
Slice and Dice P_DESCRIPT Products Sold AvgPriceV_STATE 1.25-in. metal screw, 253$6.99TN 7.25-in. pwr. saw blade8$14.99KY B&D cordless drill, 1/2- in. 1$38.95FL B&D jigsaw, 12-in. blade 1$109.92TN Claw hammer5$9.95TN Hicut chain saw, 16 in.1$256.99TN Hrd. cloth, 1/4-in., 2x501$39.95GA Rat-tail file, 1/8-in. fine6$4.99KY Steel matting, 4'x8'x1/6",.5" mesh 3$119.95FL Viewing Data by Vendor state
Data-driven DSS Permits users to drill-down from summary to detail Permits users to drill-up from details to summary Permits users to view data on different dimensions
Data warehouse Database Designed for decision support Batch updated Large amounts of data Can be subject oriented (Data Mart) Uses consistent definitions and formats Time-variant Historical data Not the same as OLAP
Sources of data for Data Warehouse From relational databases or Non-relational data sources Integration of data from distributed and differently structured databases Definitions and formats may not be consistent across sources Physical Separation of data used in daily operations from data used for purposes of reporting, decision support, analysis and controlling.
Data warehouse ETL = Extract, Transform and Load
Multi-dimensional analysis or OLAP systems
Multi-dimensional analysis OLAP software for creating multi- dimensional representations Original data comes from normalized two- dimensional tables Example database Tables: Vendors, Products, Outlets, Sales persons, sale records
Example application Vendors supply products Products belong to product-lines Products sold by specific sales persons Products sold at specific outlets Outlets located in specific regions Sales done in specific time period Can you spot the dimensions of sales revenue data?
Dimensional view of Sales data
How would you “Drill down” and “Slice and Dice”
Executive Information Systems
EIS To support information needs of the executive Data for specific issues and problems Can do trend analysis, and drill down Uses internal and external data Dashboard applications
Differences between data that is available and data that is required FactorsOperating DataDSS Data Data Structuresnormalizedintegrated Time Spancurrenthistorical Summarizationnoneextensive in some systems Data Volatilityvolatilenon-volatile Data Dimensionsone dimensionmultiple dimensions Metadatadesirablerequired and important
Normalized data structure to Integrated data structure Operating data Two dimensional flat tables – suitable for quick update, insert and delete operations Efficient usage of hard drive space Derived and derivable data not stored Integrated data Denormalized; for quick retrieval Frequently used summary and derived data is created and stored