A Critique of the Proposed National Education Policy Reform
Business Model Ontologies in OLAP Cubes
1. CAiSE’13CAiSE 13
Business Model Ontologies
in OLAP Cubes
Christoph Schütz, Bernd Neumayr, Michael Schreflp , y ,
This work was supported by the FIT-IT research program of the Austrian Federal Ministry for Transport,
Innovation, and Technology under grant FFG-829594 for the Semantic Cockpit project.
2. CAiSE’13
Overview
CAiSE 13
Introduction
■ Facts with Ontology-valued Measures
□ Base Facts
□ Shared Facts
■ OLAP with Ontology-valued Measures
M□ Merge
□ Abstraction
Implementation■ Implementation
■ Summary and Future Work
2JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
3. CAiSE’13
Introduction
CAiSE 13
■ Traditional cube: Numeric measures
■ Many real-world facts do not boil down to numeric values
■ How do you measure complex situations?
Example: intensity of competition
3JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
4. CAiSE’13
Introduction
CAiSE 13
■ Analysts compile strategic analysis documents
■ Not (only) numeric measures■ Not (only) numeric measures
Ontology-valued measures
x:Marketing/Germany/Q1-
2012
x:Marketing/France/Q1-2012
x:Germany/Sales/Q2-2012
Germany
x:Germany/Production/Q2-2012
x:MegaCar x:sells x:MegaSUV
x:Familiesx:hasClient
x:We
x:Our_Truck
x:produces
x:Development/Germany/
Q1-2012
x:Development/France/Q1-
2012
x:France/Sales/Q2-2012 x:France/Production/Q2-2012
x:We x:sells x:Our_Truck
x:Food_Incx:hasClient x:MegaCar
x:hasSupplier
x:France/Sales/Q2 2012 x:France/Production/Q2 2012
Q1-2012France
x:MegaCar x:sells x:MegaSUV
x:Singlesx:hasClient
W ll O r SUV
x:MegaCar
x:produces
x:MegaSUV
x:MidiCarx:hasSupplier
x:hasSupplier
4JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge EngineeringSales Production
Q2-2012
x:We x:sells x:Our_SUV
x:Familiesx:hasClient
x:We
x:Our_SUV
x:produces
asSupp e
5. CAiSE’13
Introduction
CAiSE 13
■ Roll-up along the dimension hierarchies
Combine knowledge from different contexts■ Combine knowledge from different contexts
U
nion
Intersection
5JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
6. CAiSE’13
Introduction
CAiSE 13
■ Roll-up along the dimension hierarchies
Combine knowledge from different contexts■ Combine knowledge from different contexts
U
nion
Intersection
6JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
7. CAiSE’13
Introduction
CAiSE 13
■ Roll-up along the dimension hierarchies
Combine knowledge from different contexts■ Combine knowledge from different contexts
U
nion
Intersection
7JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
8. CAiSE’13
Introduction
CAiSE 13
■ Roll-up along the dimension hierarchies
Combine knowledge from different contexts■ Combine knowledge from different contexts
U
nion
Intersection
8JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
9. CAiSE’13
Introduction
CAiSE 13
■ Roll-up along the dimension hierarchies
Combine knowledge from different contexts■ Combine knowledge from different contexts
U
nion
Intersection
9JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
10. CAiSE’13
Introduction
CAiSE 13
■ Roll-up along the dimension hierarchies
Combine knowledge from different contexts■ Combine knowledge from different contexts
U
nion
Intersection
10JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
11. CAiSE’13
Introduction
CAiSE 13
■ Roll-up along the dimension hierarchies
Combine knowledge from different contexts■ Combine knowledge from different contexts
U
nion
Intersection
11JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
12. CAiSE’13
Introduction
CAiSE 13
■ Alter the granularity of the ontologies
Use knowledge from the ontologies for■ Use knowledge from the ontologies for
abstraction
x:Europe/Sales/Q2-2012 x:Europe/Sales/Q2-2012
x:MegaCar x:sells x:SUVsx:MegaCar x:sells x:MegaSUV
h Cli
Europe Abstract Europe
x:Singlesx:hasClient
x:Familiesx:hasClient
x:hasClient x:Food_Inc
x:Households
x:Corporate
x:hasClient
x:hasClient
x:sells
Q2-2012
x:We
x:Our_Truck
x:sells
x:Our_SUV
x:sells
x:hasClient
x:Trucksx:We
x:hasClient
x:sells
12JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
Sales
Q2-2012
Sales
x:sells
13. CAiSE’13
Introduction
CAiSE 13
■ Alter the granularity of the ontologies
Use knowledge from the ontologies for■ Use knowledge from the ontologies for
abstraction
x:Europe/Sales/Q2-2012 x:Europe/Sales/Q2-2012
x:MegaCar x:sells x:SUVsx:MegaCar x:sells x:MegaSUV
h Cli
Europe Abstract Europe
x:Singlesx:hasClient
x:Familiesx:hasClient
x:hasClient x:Food_Inc
x:Households
x:Corporate
x:hasClient
x:hasClient
x:sells
Q2-2012
x:We
x:Our_Truck
x:sells
x:Our_SUV
x:sells
x:hasClient
x:Trucksx:We
x:hasClient
x:sells
13JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
Sales
Q2-2012
Sales
x:sells
14. CAiSE’13
Introduction
CAiSE 13
■ Alter the granularity of the ontologies
Use knowledge from the ontologies for■ Use knowledge from the ontologies for
abstraction
x:Europe/Sales/Q2-2012 x:Europe/Sales/Q2-2012
x:MegaCar x:sells x:SUVsx:MegaCar x:sells x:MegaSUV
h Cli
Europe Abstract Europe
x:Singlesx:hasClient
x:Familiesx:hasClient
x:hasClient x:Food_Inc
x:Households
x:Corporate
x:hasClient
x:hasClient
x:sells
Q2-2012
x:We
x:Our_Truck
x:sells
x:Our_SUV
x:sells
x:hasClient
x:Trucksx:We
x:hasClient
x:sells
14JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
Sales
Q2-2012
Sales
x:sells
15. CAiSE’13
Overview
CAiSE 13
■ Introduction
Facts with Ontology-valued Measures
□ Base Facts
□ Shared Facts
■ OLAP with Ontology-valued Measures
M□ Merge
□ Abstraction
Implementation■ Implementation
■ Summary and Future Work
15JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
24. CAiSE’13
Shared Facts
CAiSE 13
■ Shared facts represent asserted knowledge at more abstract
l l f b t tilevels of abstraction
■ Base facts inherit knowledge represented in the more abstract
shared factsshared facts
■ Shared facts facilitate the analysis
‹ all › ‹ all ›
‹ all ›Location
Organization
Time
‹ continent ›
‹ country ›
‹ year ›
‹ quarter ›
‹ department ›
Strategy
24JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
+ competition: RDF
Strategy
25. CAiSE’13
Shared Facts
CAiSE 13
x:Organization Model
Time: ‹ all ›
: Strategy
Organization: ‹ all ›
x:Organization_Model
Location: ‹ all ›
x:OurTruck
x:ProductModel
x:Enterprise
rea:Agent
rdfs:subClassOf
rdf:type
rdf:type
+ competition =
x:Organization_Model
S l M d l
rdf:type
x:We
x:OurTruckx:Enterprise
x:FunnyCar x:CleverCar
rdf:type rdf:type x:OurSUV
Sales: ‹ department ›
Organization: ‹ all ›
x:Sales_Model
x:Families x:Singles
x:Households
x:OurTruck x:OurSUV
x:SUVsx:Trucks
rea:grouping
rea:grouping
rea:groupingrea:grouping
Time: ‹ all ›
+ competition =
x:Sales_Model
: Strategy
Location: ‹ all ›
x:Families
rea:Group
x:Singles
rdf:typerdf:type
x:Enterprise
x:Food_Inc
rdf:type
x:PaymentType
x:Money
rdf:type
Production: ‹ department ›
Organization: ‹ all ›
x:Production_Model
x:ProductModel
rdf:type
x:ToolModel
rdf:type
25JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
Location: ‹ all › Time: ‹ all ›
+ competition =
x:Production_Model
: Strategy
x:Enterprise x:Binford
rdf:type
x:OurTruckEngine
x:CleverCarChassis
x:FunnySUVEngine
rdf:type
rdf:type
x:BinfordRobot
26. CAiSE’13
Shared Facts
CAiSE 13
x:Organization Model
Time: ‹ all ›
: Strategy
Organization: ‹ all ›
x:Organization_Model
Location: ‹ all ›
x:OurTruck
x:ProductModel
x:Enterprise
rea:Agent
rdfs:subClassOf
rdf:type
rdf:type
+ metamodel
and
common
+ competition =
x:Organization_Model
S l M d l
rdf:type
x:We
x:OurTruckx:Enterprise
x:FunnyCar x:CleverCar
rdf:type rdf:type x:OurSUV
application
model
Sales: ‹ department ›
Organization: ‹ all ›
x:Sales_Model
x:Families x:Singles
x:Households
x:OurTruck x:OurSUV
x:SUVsx:Trucks
rea:grouping
rea:grouping
rea:groupingrea:grouping
+ salesTime: ‹ all ›
+ competition =
x:Sales_Model
: Strategy
Location: ‹ all ›
x:Families
rea:Group
x:Singles
rdf:typerdf:type
x:Enterprise
x:Food_Inc
rdf:type
x:PaymentType
x:Money
rdf:type
+ sales
application
model
Production: ‹ department ›
Organization: ‹ all ›
x:Production_Model
x:ProductModel
rdf:type
x:ToolModel
rdf:type
+ production
26JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
Location: ‹ all › Time: ‹ all ›
+ competition =
x:Production_Model
: Strategy
x:Enterprise x:Binford
rdf:type
x:OurTruckEngine
x:CleverCarChassis
x:FunnySUVEngine
rdf:type
rdf:type
x:BinfordRobot
+ production
application
model
41. CAiSE’13
Overview
CAiSE 13
■ Introduction
■ Facts with Ontology-valued Measures
□ Base Facts
□ Shared Facts
■ OLAP with Ontology-valued Measures
M□ Merge
□ Abstraction
Implementation Implementation
■ Summary and Future Work
41JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
42. CAiSE’13
Implementation
CAiSE 13
■ Based on hetero-homogeneous data warehouse
htt //hh d dk i li t/http://hh-dw.dke.uni-linz.ac.at/
O l DB f th ltidi i l d l■ Oracle DB for the multidimensional model
■ Jena tuple store for RDF graphs and Jena framework for
SPARQL queriesSPARQL queries
U i th ltidi i l d l i O l DB i d f■ Using the multidimensional model in Oracle DB as index for
calculating the inherited knowledge
42JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
43. CAiSE’13
Overview
CAiSE 13
■ Introduction
■ Facts with Ontology-valued Measures
□ Base Facts
□ Shared Facts
■ OLAP with Ontology-valued Measures
M□ Merge
□ Abstraction
Implementation■ Implementation
Summary and Future Work
43JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering
44. CAiSE’13
Summary and Future Work
CAiSE 13
■ Ontology-valued measures for complex real-world facts that do
t b il d t inot boil down to a numeric measure
Oth b i d l t l i f t l l d■ Other business model ontologies for ontology-valued measures
In particular: e3value and its variants, e.g., e3forces
■ Provide for easier querying, examine other query languages
44JKU Linz Institut für Wirtschaftsinformatik – Data & Knowledge Engineering