XBRL Taxonomy Engineering

1
XBRL Taxonomy Engineering
Maciej Piechocki IASC Foundation XBRL Team,
UK Freiberg University of Technology, Germany

• Munich, 5 June 2007

2
Engineering Approach vs. Artistic Approach

• Transparent/ repeatable
• Workload divisiable
• Structured
• Systematic
• Scalable

• Usually individual instead of team output
• Usually not justifiable
• Usually not repeatable
• Usually hardly explainable

3
Agenda

• Introduction
• Software, Knowledge and Ontology Engineering
• XBRL Taxonomy Engineering
• XBRL Taxonomy Lifecycle
• XBRL Taxonomy Development Process Model
• Conclusions

4
Introduction

• Issues
• IFRS taxonomy for reporting from January, 1st
  2006 was released only on August, 15th 2006
• German accounting principles taxonomy update from
  XBRL 2.0 to the XBRL 2.1 specification (released
  on 31 December 2003) is still under development
• FINREP taxonomy of the Central European Banking
  Supervisors based on the regulations being in
  place at the beginning of 2006 was released in
  September 2006
• For all released taxonomies (IFRS, US-GAAP,
  COREP, FINREP, GermanAP, XBRL GL etc.) number of
  bugs are reported

5
Introduction

• Conclusions
• Lack of guidelines or documented experiences on
  taxonomy development process
• Systematic, formal and generic approach is
  necessary for taxonomy development process
• Software engineering, ontology engineering and
  knowledge engineering offer such approaches
• Taxonomy engineering as taxonomy development
  process model needs to be defined

6
Agenda

• Introduction
• Software, Knowledge and Ontology Engineering
• XBRL Taxonomy Engineering
• XBRL Taxonomy Lifecycle
• XBRL Taxonomy Development Process Model
• Conclusions

7
Software, Knowledge and Ontology Engineering
Software engineering Knowledge engineering Ontology engineering
is the application of a systematic, disciplined, quantifiable approach to the development, operation, and maintenance of software. (IEEE 1990) refers to the building, maintaining and development of knowledge-based systems. Knowledge engineering understands knowledge that is produced by socio-cognitive aggregates (mainly humans) and is structured according to our understanding of how human reasoning and logic works. (Kraaijenbrink 2006) is a research methodology which gives us design rationale of a knowledge base, kernel conceptualization of the world of interest, strict definition of basic meanings of basic concepts together with sophisticated theories and technologies enabling accumulation of knowledge which is dispensable for modeling the real world. (Mizoguchi and Ikeda 1996)
8
Software, Knowledge and Ontology Engineering
Taxonomy Engineering
9
Software, Knowledge and Ontology Engineering
Ontology Engineering
Software Engineering
Water- fall model
Mod. Waterf. model
Methon- tology
Generic
eXtreme Progr.
Proto- typing
On-to Know- ledge
Common KADS
SENSUS
Specific
TOVE
KACTUS
10
Agenda

• Introduction
• Software, Knowledge and Ontology Engineering
• XBRL Taxonomy Engineering
• XBRL Taxonomy Lifecycle
• XBRL Taxonomy Development Process Model
• Conclusions

11
XBRL Taxonomy Lifecycle
12
XBRL Taxonomy Lifecycle

• Planning and analysis phase
• Taxonomy development project requirements
• Determination of taxonomy scope
• Choice of the core taxonomy
• Design of the taxonomy framework
• Determination of the legal background for the
  taxonomy
• Result taxonomy requirements documentation

13
XBRL Taxonomy Lifecycle

• Taxonomy design phase
• Mixture of domain and technical experts
• Modeling the reality
• Lack of a common modeling approach in XBRL and a
  formal notation
• Use of spreadsheets to represent the knowledge
  base
• Result information model

14
XBRL Taxonomy Lifecycle

• Taxonomy building phase
• Transfer of the knowledge base from the
  information model into the coded structures of
  the XBRL taxonomy
• Technical knowledge necessary
• FRTA with established rules and conventions
• Taxonomy patterns which define the possible
  structures appearing in financial reports and
  their representation in taxonomy code
• Initial taxonomy working drafts are usually not
  using the full functionality of XBRL
• Testing of the taxonomy working drafts at the end
  of each development cycle
• Result taxonomy working drafts

15
XBRL Taxonomy Lifecycle

• Taxonomy testing phase
• Various test levels
• Basic prerequisite is the compliance of the
  taxonomy with XML, XML Schema, XLink
  specifications, and XBRL specification
• FRTA rules should be obeyed for the constructed
  taxonomies in order to follow the best practices
  of XBRL taxonomy development
• Domain experts review of the final working draft
• Samples and real instance documents
• Taxonomy users involvement
• Result taxonomy exposure draft

16
XBRL Taxonomy Lifecycle

• Taxonomy publication and recognition phase
• Publication on the websites of the corresponding
  organization as well as they announcement on the
  website of XBRL International
• Recognition process at the XBRL International
  level assigning to a taxonomy either the status
  acknowledged or approved
• XBRL International can approve, ask for changes
  or reject the taxonomy during the recognition
  process
• Result taxonomy final version

17
XBRL Taxonomy Lifecycle

• Taxonomy usage and maintenance phase
• Reporting entities and also software vendors
  building in the XBRL reporting capabilities into
  their software products
• Taxonomy developers monitor the taxonomy usage
  and track the issues and bugs the users
  experience with the taxonomy
• Feedback has to be incorporated in the next
  version of the information model and has to be
  considered in the next taxonomy release
• Problematic of taxonomy versioning

18
XBRL Taxonomy Development Process Model
19
XBRL Taxonomy Engineering
Criteria Software Engineering Ontology Engineering Taxonomy Engineering
Generic model Yes No Yes
Phase oriented models Yes No Yes
Cycles allowed Yes/ No Yes Yes
Knowledge acquisition phase No Yes Yes
Extensive test phase Yes No Yes
Reusing of existing software product/ ontology Not addressed Yes Yes
Domain experts cooperation Not addressed Yes Yes
Result Software product Ontology or knowledge-based system XBRL taxonomy
20
Agenda

• Introduction
• Software, Knowledge and Ontology Engineering
• XBRL Taxonomy Engineering
• XBRL Taxonomy Lifecycle
• XBRL Taxonomy Development Process Model
• Conclusions

21
Conclusions

• Model fulfilling the IEEE-norm 1074-1997
• Definition derived from IEEE Standard 610.12
• Taxonomy engineering is the application of
  systematic, formal, quantifiable approach to the
  design, building, usage and maintenance of XBRL
  taxonomies, that is application of engineering to
  XBRL taxonomies.
• Further directions
• Enhance the taxonomy engineering with detailed
  steps and subsequent tools and modelling
  techniques
• Consider agile software development techniques

22
Questions

• More info
• www.iasb.org/xbrl
• www.wiwi.tu-freiberg.de/wi/infolog

Maciej Piechocki IASC Foundation XBRL Team,
UK Freiberg University of Technology, Germany