Logical design DATA WAREHOUSE: DESIGN Logical design. We address the relational model (ROLAP)

Transcription

1 atabase and ata Mining Group of atabase and ata Mining Group of B MG ata warehouse design atabase and ata Mining Group of atabase and data mining group, M B G Logical design ATA WAREHOUSE: ESIGN - 37 Logical design We address the relational model (ROLAP) inputs conceptual fact schema workload data volume system constraints output relational logical schema Based on different principles with respect to traditional logical design data redundancy table denormalization atabase and data mining group, ATA WAREHOUSE: ESIGN - 38 Pag. 1

2 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design Star schema imensions one table for each dimension surrogate (generated) primary key it contains all dimension attributes hierarchies are not explicitly represented all attributes in a table are at the same level totally denormalized representation it causes data redundancy Facts one fact table for each fact schema primary key composed by foreign keys of all dimensions measures are attributes of the fact table atabase and data mining group, ATA WAREHOUSE: ESIGN - 39 Star schema atabase and data mining group, Supplier Type Category Salesman imension table imension table Week Week_I Week Month _I Type Category Supplier Month Week SALES Amount Shop_I Week_I _I Amount Fact table Shop City Country Shop Shop_I Shop City Country Salesman imension table From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 40 Pag. 2

3 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design Star schema atabase and data mining group, Shop_I Shop City Country Salesman 1 N1 RM I R1 2 N2 RM I R1 3 N3 MI I R2 4 N4 MI I R2 imension Table Shop_I Week_I _I Amount Fact Table Week_I Week Month 1 Jan1 Jan. 2 Jan2 Jan. 3 Feb1 Feb. 4 Feb2 Feb. imension Table _I Type Category Supplier 1 P1 A X F1 2 P2 A X F1 3 P3 B X F2 4 P4 B X F2 From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 41 Snowflake schema atabase and data mining group, Some functional dependencies are separated, by partitioning dimension data in several tables a new table separates two branches of a dimensional hierarchy (hierarchy is cut on a given attribute) a new foreign key correlates the dimension with the new table ecrease in space required for storing the dimension decrease is frequently not significant Increase in cost for reading entire dimension one or more joins are needed ATA WAREHOUSE: ESIGN - 42 Pag. 3

4 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design Snowflake schema Type Supplier Category atabase and data mining group, Salesman Week Week_I Week Month _I Type_I Supplier Type Type_I Type Category Month Week SALES Amount Shop_I Week_I _I Amount Shop City Country Shop Shop_I Shop City_I Salesman City City_I City Country From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 43 Foreign key Snowflake schema atabase and data mining group, Type_I Type Category 1 A X 2 B X _I Supplier Type_I 1 P1 F1 1 2 P2 F1 1 3 P3 F2 2 4 P4 F2 2 Week_I. Week Month 1 Jan1 Jan. 2 Jan2 Jan. 3 Feb1 Feb. 4 Feb2 Feb. Shop_I Week_I _I Amount Shop_I Shop City_I Salesman 1 N1 1 R1 2 N2 1 R1 3 N3 2 R2 4 N4 2 R2 City_I City Country 1 RM I 2 MI I From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 44 Pag. 4

5 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design Star or snowflake? The snowflake schema is usually not recommended atabase and data mining group, storage space decrease is rarely beneficial most storage space is consumed by the fact table (difference with dimensions is several orders of magnitude) cost of join execution may be significant The snowflake schema may be useful when part of a hierarchy is shared among dimensions (e.g., geographic hierarchy) for materialized views, which require an aggregate representation of the corresponding dimensions ATA WAREHOUSE: ESIGN - 45 Multiple edges atabase and data mining group, genre SALE author book quantity income date month year Implementation techniques bridge table new table which models many to many relationship new attribute weighting the contribution of tuples in the relationship push down multiple edge integrated in the fact table new corresponding dimension in the fact table ATA WAREHOUSE: ESIGN - 46 Pag. 5

6 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design Multiple edges atabase and data mining group, genre SALE author book quantity income date month year Sales Book_I ate_i Income Books Book_I Book Genre Authors Author_I Author BRIGE Book_I Author_I Weight Sales Book_I Author_I ate_i Income Books Book_I Book Genre Authors Author_I Author From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 47 Multiple edges atabase and data mining group, Queries Weighted query: consider the weight of the multiple edge example: author income by using bridge table: SUM(Income*weight) group by I_author Impact query: do not consider the weight of the multiple edge example: book copies sold for each author by using bridge table: SUM() group by I_author ATA WAREHOUSE: ESIGN - 48 Pag. 6

7 atabase and ata Mining Group of atabase and ata Mining Group of B MG ata warehouse design atabase and ata Mining Group of Multiple edges atabase and data mining group, Comparison weight is explicited in the bridge table, but wired in the fact table for push down (push down) hard to perform impact queries (push down) weight is computed when feeding the W (push down) weight modifications are hard push down causes significant redundancy in the fact table query execution time for push down less joins higher cardinality for the fact table ATA WAREHOUSE: ESIGN - 49 egenerate dimensions atabase and data mining group, imensions with a single attribute Category Supplier Type Shipping Mode ORER LINE Return code Amount Line Order Status Order Customer City ATA WAREHOUSE: ESIGN - 50 Pag. 7

8 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design egenerate dimensions atabase and data mining group, Implementations (usually) directly integrated into the fact table only for attributes with a (very) small size junk dimension single dimension containing several degenerate dimensions no functional dependencies among attributes in the junk dimension all attribute value combinations are allowed feasible only for attribute domains with small cardinality ATA WAREHOUSE: ESIGN - 51 Junk dimension atabase and data mining group, Order Line Order_I _I SRL_I Amount Order Order_I Order Customer City_I SRL SRL_I ShippingMode ReturnCode LineOrderStatus From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 52 Pag. 8

9 atabase and ata Mining Group of atabase and ata Mining Group of B MG ata warehouse design atabase and ata Mining Group of Materialized views atabase and data mining group, Precomputed summaries for the fact table explicitly stored in the data warehouse provide a performance increase for aggregate queries v 1 = {product, date, shop} v 2 = {type, date, city} v 4 = {type, month, region} v 3 = {category, month, city} v 5 = {quarter, region} From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 53 Materialized views atabase and data mining group, Materialized views may be exploited for answering several different queries not for all aggregation operators {a,c} d b c a {b,c} {a,d} {c} {b,d} {a} {d} {b} { } Multidimensional lattice From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 54 Pag. 9

10 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design atabase and data mining group, Materialized view selection Huge number of allowed aggregations most attribute combinations are eligible Selection of the best materialized view set Cost function minimization query execution cost view maintainance (update) cost Constraints available space time window for update response time data freshness ATA WAREHOUSE: ESIGN - 55 atabase and data mining group, Materialized view selection {a,c} {b,c} {a,d} {c} {b,d} {a} q 3 q 1 {d} { } {b} q 2 + Multidimensional lattice = candidate views, possibly useful to increase workload query performance From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 56 Pag. 10

11 atabase and ata Mining Group of atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design atabase and data mining group, Materialized view selection {a,c} Cost minimization {c} {b,c} {b,d} {a,d} {a} q 3 isk space Update window Query cost {d} {b} q 1 q 2 { } From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 57 atabase and data mining group, Materialized view selection {a,c} Space and time minimization {b,c} {a,d} {c} {b,d} {a} q 3 isk space Update window Query cost q 1 {d} {b} q 2 { } From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 58 Pag. 11

12 atabase and ata Mining Group of B MG atabase and ata Mining Group of ata warehouse design atabase and data mining group, Materialized view selection {a,c} All constraints {b,c} {a,d} {c} {b,d} {a} q 3 isk space Update window Query cost {d} {b} q 1 q 2 { } From Golfarelli, Rizzi, ata warehouse, teoria e pratica della progettazione, McGraw Hill 2006 ATA WAREHOUSE: ESIGN - 59 Pag. 12