Data Warehouses Incremental Integration of The Hetero-Homogeneous Approach Christoph Schütz, Michael Schrefl Bernd Neumayr, Daniel Sierninger.

Data WarehousesIncremental Integration of

The Hetero-Homogeneous Approach

Christoph Schütz, Michael SchreflBernd Neumayr, Daniel Sierninger

2JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Overview

■ Motivation

■ Hetero-Homogeneous Data Warehouses□ Modeling Hetero-Homogeneous Dimensions

□ Modeling Hetero-Homogeneous Cubes

■ Incremental Integration□ Integrating Hetero-Homogeneous Dimensions

□ Integrating Hetero-Homogeneous Cubes

■ Prototype

3JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

MOTIVATIONHeterogeneities in Conceptual Modeling

4JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

listPrice = 80color = 'Red'length = 50

Heterogeneities in Object-oriented Design

ProductModel

listPriceauthorcolorlength

DaVinciCode :ProductModel

listPrice = 10author = 'Brown'color = NULLlength = NULL

HonoluluSkirt:ProductModel

listPrice = 80author = NULLcolor = 'Red'length = 50

ProductModel

listPrice

BookTitle

author

ClothingModel

colorlength

listPrice = 10author = 'Brown‘

5JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Heterogeneities in Data Warehouse Modeling

HeterogeneousHomogeneous

6JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Hetero-Homogeneous Dimensions

7JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

HETERO-HOMOGENEOUSDATA WAREHOUSES

Modeling Hetero-Homogeneous Dimensions and Cubes

8JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Modeling Hetero-Homogeneous Dimensions

Multilevel Object (M-Object)

9JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Modeling Hetero-Homogeneous Cubes

Multilevel Relationship(M-Relationship)

Multilevel Fact(M-Fact)

10JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

INCREMENTAL INTEGRATION

Integrating Hetero-Homogeneous Dimensions and Cubes

11JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Integrating Hetero-Homogeneous Dimensions

12JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

■ Generalization Guideline: If two m-objects share a level or attribute then they must have a common ancestor which introduces the respective level or attribute.

■ Mapping Rules: Explicit mapping rules are employed to bridge name differences and assert semantic equivalence (or difference)

Integrating Hetero-Homogeneous Dimensions

13JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Integrating Hetero-Homogeneous Dimensions

Abstract Superclass Rule

14JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Integrating Hetero-Homogeneous Cubes

15JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Integrating Hetero-Homogeneous Cubes

Set-Valued Coordinates

16JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Prototype

■ Oracle 11g and PL/SQL

■ Provides management and query functionality form-objects and m-relationships

■ The conceptual model is stored using Oracle‘s object-relational capabilities.

■ The logical model is a variant of the fact constellation schema.

■ Proof-of-concept

■ Some features still under development

17JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering

Bibliography

■ [hh-dw] Hetero-Homogeneous Data Warehouses (HH-DW). Project Homepage. http://www.hh-dw.dke.uni-linz.ac.at/

■ [NST10] Neumayr, B., Schrefl, M., Thalheim, B.: Hetero-Homogeneous Hierarchies in Data Warehouses. APCCM ’10.

■ [Sch10] Schütz, C.: Extending data warehouses with hetero-homogeneous dimension hierarchies and cubes: A proof-of-concept prototype in Oracle. Master‘s Thesis. 2010.

■ [Sie11] Sierninger, D.: Integration von Data Marts in ein globales Data Warehouse mit hetero-homogenem Schema. Master‘s Thesis. 2011.