XBRL Multidimensional Modelling

1
XBRL Multidimensional Modelling

• Maciej Piechocki
• IASC Foundation XBRL Team, UK
• Freiberg University of Technology, Germany
• Michal Piechocki
• IASC Foundation XBRL Team, UK
• Business Reporting Advisory Group, Poland

• Munich, 5 June 2007

2
Agenda

• Introduction
• Traditional Multidimensional Modelling Approaches
• XBRL Dimensions
• Conclusions

3
Multidimensionality
4
Problem Domain
5
Semantic in XBRL Data Models
6
Agenda

• Introduction
• Traditional Multidimensional Modelling Approaches
• XBRL Dimensions
• Conclusions

7
Traditional Multidimensional Modelling Approaches
8
ME/R Modell
9
ADAPT
10
Star Schema
11
Agenda

• Introduction
• Traditional Multidimensional Modelling Approaches
• XBRL Dimensions
• Conclusions

12
Basic Concepts
primary items, measures
basic financial/accounting concepts, for which
there could be a dimensional context information
created
dimensions
elements describing dimensional breakdowns
domains and domain members
elements defining details of dimensional
breakdowns
hypercubes
element grouping dimensions that may be used for
a measure (or are prohibited)
13
Modularity of XBRL Dimensions
primary taxonomy
defines measures and their relationships
template taxonomy
element grouping dimensions that may be used for
a measure (or are prohibited)
dimensional/domain-member taxonomy
contains definitions of dimensions and
domain-members and describes relations between
them
14
Semantic of Dimensional Relationships
domain-member

• defines hierarchical relations for measures
  (additionally, children elements inherit
  dimensional information described for their
  parents)
• defines hierarchical relations of domain members

all
connects measure to a hypercube implying use of
dimensions (attached to hypercube)
notAll
connects measure to a hypercube prohibiting use
of dimensions (attached to hypercube)
hypercube-dimension
connects hypercube and dimension
dimension-domain
connects dimension to its member (usually top
level domain)
dimension-default
connects dimension to its default member (usually
top level domain)
15
measure
typed dimension
domain-member targetRoleanyURI
substitutionGroup"xbrliitem"
substitutionGroup"xbrldtdimensionItem" abstract
"true" typedDomainRefanyURI
all, notAll closedboolean contextElement
segment, scenario targetRoleanyURI
typed dimension domain
non XBRL element defining restriction or complex
model restricting content of typed dimension
hypercube
substitutionGroup"xbrldthypercubeItem" abstract
"true"
hypercube-dimension targetRoleanyURI
explicit dimension
dimension-domain usableboolean targetRoleanyUR
I
substitutionGroup"xbrldtdimensionItem" abstract
"true"
domain or domain member
domain-member usableboolean targetRoleanyURI
dimension-default
substitutionGroup"xbrliitem" abstract"true"
domain or domain member
domain-member usableboolean targetRoleanyURI
substitutionGroup"xbrliitem" abstract"true"
domain-member usableboolean targetRoleanyURI
inheritance of dimensional information from
parents
16
Conclusions
17
Agenda

• Introduction
• Traditional Multidimensional Modelling Approaches
• XBRL Dimensions
• Conclusions

18
Conclusions

• Two different modelling approaches
• Different goals and requirements of the
  approaches
• XBRL modelling semantically stronger
• Interoperability issues
• Possible solutions
• Flattening of dimensional XBRL
• Translations of data models
• ETL and inverted ETL processing
• ...
• Still theoretical XBRL as data warehouse

19
Questions

• More info
• www.wiwi.tu-freiberg.de/wi/infolog
• www.br-ag.eu

Maciej Piechocki IASC Foundation XBRL Team,
UK Freiberg University of Technology,
Germany Michal Piechocki IASC Foundation XBRL
Team, UK Business Reporting Advisory Group,
Poland