XBRL AND THE MULTIDIMENSIONAL DATA MODEL
Ignacio Santos and Elena Castro
Carlos III University of Madrid, Computer Science Department
Avda. de la Universidad Nº 30, 28911 Leganés, Madrid, Spain
Keywords: XBRL (eXtensible Business Reporting Language), XML (eXtensible Markup Language) Taxonomy, XDT
(XBRL Dimensional Taxonomy), Conceptual data model, MDM (Multidimensional data Model).
Abstract: Over the past ten years, there has been a significant increasing of the development of XML and Data
Warehouse (DW) applications, and, in turn, more and more applications need to interact, and with different
technologies. In parallel, the economic data in the last ten years have also evolved, increasingly companies
and financial institutions need more information, in addition, this information must be reliable and on time.
Nowadays, it is taking a significant rise for XBRL standard, based on XML. This language is mainly used in
accounting reports and this consists of a set of taxonomies, which define different accounting regulations of
a specific report. XBRL is becoming a global de facto standard. XBRL reports are created from various
sources and are validated at source, so that, this is syntactically correct. XBRL represents business
information, and it is multidimensional, and therefore the logical destination is a DW. This paper aims to
analyze the data model of XBRL and its semantics, and how to map this data model to the Multidimensional
Data Model (Conceptual Model) and in turn to the Logical Model, either ROLAP (Relational OLAP),
MOLAP (Multidimensional OLAP), or HOLAP (Hybrid OLAP), so they can be analyzed by business users.
1 INTRODUCTION
The accounting world is static and dynamic. It is
static, because the accounting reports use strict
regulations, such as the International Financial
Reporting Standard (IFRS), or the Accounting
General Plan in Spain. For example, the financial
statements of credit institutions are specific reports,
and they are defined with one or more taxonomies,
but they have structures and meaning fixed. In
addition, the accounting world is dynamic because
accounting directives vary with time, turning up
versioning problems, or location. For example,
accounting regulations on capital movements are
based on Basel II (Basel II, 2004), but because
recent financial crisis, these regulations will be
amended in Basel III (Basel III, 2010). All of these
reports entail that a large number of companies,
economic and financial entities, must be called upon
to use a large number of resources in Information
Technology (IT). This paper presents the basis of
study to analyze and transform the data model of this
set of taxonomies in an automated way on a
multidimensional data model.
In the section two we develop the research work
about XML and the Multidimensional Model. The
section three, describes the XBRL data model. In the
section four, we analyze our proposal, about the
transformation of the models. The section five, an
example is shown. And, in the last section, some
conclusions and ongoing work are presented.
2 FROM XML TO THE
MULTIDIMENSIONAL MODEL
In this section, we check some works about XML,
and Data Warehouse and its connections.
In Kurtev et al., XML documents are
transformed into other formats, using XML Schemas
(Kurtev, 2003). Ogbuje does a comparison and
mapping of XML and Resource Description
Framework technologies (RDF), and with the Model
Driven Architecture (MDA) architecture (Ogbuje,
2002). XML documents are usually stored in
database. XML was not designed for data storage,
but it is a data description language (Lucas-Torres,
2009). It is possible to transform of an automatic
way the multidimensional model to a logical model,
using MDA (Mazón, 2007), (Pardillo, 2008),
(Mazón, 2005), (Castro, 2004), (Golfarelli, 2001).
161
Santos I. and Castro E..
XBRL AND THE MULTIDIMENSIONAL DATA MODEL.
DOI: 10.5220/0003399001610164
In Proceedings of the 7th International Conference on Web Information Systems and Technologies (WEBIST-2011), pages 161-164
ISBN: 978-989-8425-51-5
Copyright
c
2011 SCITEPRESS (Science and Technology Publications, Lda.)
These developments are based on the Relational
Model (Logical Model), and not in the E/R model
(Conceptual Model). On the other hand, the
conversion between heterogeneous models and
applications with different sources, especially Web
applications and databases has been analyzed
(Golfarelli, 2001). It can be compared XML DTD
(Document Type Definition) model and the E/R, by
analyzing the mapping of DTD to E/R model, to
Object/Relational Model (O/R) and to Relational.
It’s possible to analyze also the entire life cycle,
from the XML document to databases (Golfarelli,
2001), (Castro, 2004). It is proposed an XML
notation of referential integrity (Pakorný, 2002).
3 XBRL REPORTING
LANGUAGE
Since the bankruptcy of Enron Corporation, was the
need to streamline the information received by the
supervisors of the stock markets to get this
information quickly, without mistakes, for security,
and also easy to analyze. XBRL is an XML-based
standard for financial reporting. Charles Hoffman,
an accountant and auditor, in April 1998 proposed to
automate the exchange of financial information,
developing a prototype of the financial statements
and audit programs, defining the bases of XBRL
(XBRL, 2001). These reports are used, for example,
to send balance sheets between Town Councils,
Autonomous Regions, and Treasury Ministry, or to
send results of corporate accounts to the Association
of Spanish Property and Commercial Registrars in
Spain, or to send financial statements to the
Supervision, and so on. XBRL is being used in most
of the world and is expanding rapidly, using it in
more than a half of European countries. Examples of
financial institutions are Bank of Spain, National
Commission of Securities Market of Spain (CNMV),
Bank of Portugal, "Deutsche Börse", Committee of
European Banking Supervisor (CEBS), in the U.S.,
Board of Governors of the Federal Reserve System
(FED), US Securities and Exchange Commission
(SEC), and other countries as Japan, India, etc. And,
in last autumn China incorporated this standard. A
financial report is based on business rules such as
IFRS, Basel II (Convention on the law and banking
regulation), and others. These standards may be
international, and/or national and/or local and/or
company, or financial institution. This set of rules
can be inherited or overwritten. XBRL allows
modelling the information, the expression and the
semantic meaning commonly required in the econo-
mic and financial reports. It uses XML syntax and
related technologies like XML Schema, XLink,
XPath and Namespaces to provide semantic
meaning. (Hernández-Ros, 2009).
Figure 1: UML design of XBRL Schemas and linkbases
(DTS).
XBRL semantic information is separated from
the application software, and it uses the XML
standard, extending its definitions to support
semantic. A report consists of an XBRL instance,
which is the report itself, and this report has a set of
XML Schemas or XBRL Schemas, called
Discoverable Taxonomy Set (DTS), which specify
the economic concepts. As it is displayed in Figure
1, each XBRL Schema may have up to eight roles
(linkbases). These roles are the definition,
presentation, label, reference, and so on. However,
each XBRL Schema may have 0 or 1 role of a given
type, and this in turn depends directly on the
existence of taxonomy, so the diamond means
combination (no aggregation), except in dimensions,
formulas, and render, which can be used for multiple
taxonomies, and a taxonomy can have 0 or more
dimensions, renders o formulas. Also, a XBRL
Schema can have one or more labels, but
conceptually is one, is the same label of the concept,
but in different languages. Also the labels depend on
the existence of taxonomy. Each XBRL Schema and
its Linkbases (roles) are taxonomies or concept
definition, which are inherited (Martín, 2006). It
begins with a definition of concepts, in general
accounting concepts for example IFRS. From this
taxonomy or concept definitions, each country or
supranational institution adapts this taxonomy to
accounting laws, adding concepts or changing them.
XBRL Dimensional Taxonomies (XDT) has
treatment of dimensions, domains and hypercubes
WEBIST 2011 - 7th International Conference on Web Information Systems and Technologies
162
(Hernández-Ros, 2010), (Schmehl, 2009), (XBRL
International, 2001). In this specification a “Primary
Item” in the XBRL data model is a fact schema in
the Multidimensional Data Model (MDM) that
represents a report.
4 XBRL MODELING, PROPOSAL
UML provides a framework to capture more
semantic with XBRL reports (Callaghan, 2002),
(Callaghan, 2006). The XBRL reports are
multidimensional (Snijders. 2005), and there is a
need to be interoperable with other applications
(Ardenne. 2005). The XBRL reports are formed of
DTS, which use dimensions with a high meaning
and semantic constraints (Piechocki, 2007), (Felden,
2007). This approach is defined by XDT
(Hernández-Ros, 2010).
Figure 2: Global transformation of the XBRL metamodel.
We propose to develop a process of
transformation of XBRL metadata, to metadata for a
multidimensional database (Figure 2), but we want
to analyze the semantic and syntactic anomalies that
the designer finds when the taxonomy is created
(Schmehl, 2009). From a report, we get the DTSs
and we will obtain an abstract data model XBRL.
Since this model, we do a transformation to MDM
(Conceptual Model). Our process with the instance,
its taxonomies (DTS), with the help of metadata
definitions of XBRL 2.1, Dimensions 1.0 and
Formula 1.0, we will obtain the Conceptual MDM.
And, from the Conceptual Model, we have to
transform to the Logical Model metadata. With
respect to the logical model we have three
techniques, MOLAP, ROLAP, and HOLAP. And
finally pass from each of the logical models to
physical model. In this physical model would be
implemented the coding to create the builders in the
corresponding managing.
5 EXAMPLE GUIDE
In this section, we analyze an example and
demonstrate how the mapping between the XBRL
data model and the MDM is possible. The example
is the 6610 report (Public Sector Consolidated
Balance report), that Spanish banks sent to the
Supervision, Bank of Spain in 2008 (BdE-FINREP,
2008), (BdE, 2008). This report is an adaptation to
the common framework agreement of the CEBS,
recently the European Parliament has reformed and
renamed this institution as European Banking
Authority (EBA).
In the Figure 3 is showed with the help of
Fujitsu’s XWand tool, our contribution about the
dimensional view of 6610 report. We can see the
change in the name in the multidimensional data
model of this report 6610. On the one hand we see
the “Primary item”, which is equivalent to the fact
schema, the distribution dimension, the dimension
attributes, fact attributes or measures, and the facts
.
Figure 3: View of Dimensional table with Xwand of
Fujitsu.
6 CONCLUSIONS
XBRL is a language XML-based that is especially
devoted to accounting reports, with a number of
advantages over other, consists of a set of
taxonomies that will be inherited to form the specific
report, to suit a particular Economic and financial
report. This makes the set of taxonomies is high, for
collecting the semantic. As we are demonstrated the
XBRL reports are dimensional and for that they can
be transformed and automated in the MDM, we want
to build a model of abstract data and analyze it,
semantically. Ongoing work with the creation the
necessary infrastructure of the automation process
XBRL AND THE MULTIDIMENSIONAL DATA MODEL
163
and its development. Finally analyze your
performance in each of the stages, and in full.
REFERENCES
Ardenne, R., December 1st, 2005. Mapping XBRL Data
into and out of conventional Data Stores. 14th XBRL
International Conference. http://14thconference.xbrl.
org/.
Basel II. June, 2004. http://www.bis.orgpubl/bcbs107.htm,
http://www.bis.org/publ/bcbsca.htm.
Basel III. September, 12
th
, 2010.http://www.bis.org/
press/p100912.htm, http://www.basel-iii-accord.com/,
http://www.oecd.org/dataoecd/42/58/45314422.pdf.
BdE. November 28th, 2008. “Descripción Nacional de la
taxonomía FINREP, Diseño Técnico. Banco de
España”. File: “es-be-finrep_tecnico_pdf”.
http://www.bde.es/fr/documentacion/taxonomias/es-
be-finrep/6-2008/2008-11-26/es-be-finrep_tecnico.pdf.
BdE-FINREP, “Taxonomía Estados financieros para
Entidades de Crédito”. November, 26
th
, 2008.”
http://www.bde.es/es/fr/documentacion/es/es-be-
finrep/6-2008/2008-11-26/index.html”.
Callaghan J. H., Nehmer R. and Sugumaran, V. August
4th, 2006. Modeling XBRL-based Applications with
UML: Developing Balanced-Scorecard Management
Appraisal Systems. AIS Electronic Library (AISel),
Proceedings of 12th Americas Conference on
Information Systems (AMCIS) 2006, Acapulco,
México, August 4-6, 2006.
Callaghan, J. H., Savage, A. and Sugumaran, V.
September 11th, 2002. Augmenting XBRL Using
UML: Improving Financial Analysis. The review of
business information Systems, 2002, Quarter 4,
volume 6, number 4. Internal title Modeling XBRL
Using UML: Improving Semantics for Financial
Analysis.
Castro Galán, E. June 14st, 2004. Una aproximación a la
reutilización de estructuras documentales definidas
con XML, en el marco de la web semántica. Thesis
PHD. Carlos III University of Madrid.
Felden, C. November 18th, 2007. Multidimensional XBRL.
DUV Deutscher Universitäts-Verlag. Book title: New
Dimensions of Business Reporting and XBRL, pages
191-209.
Golfarelli, M., Rizzi, S. and Urdoljak, B. September 3rd,
2001. Data warehouse design from XML sources.
Proceedings of the 4th ACM International workshop
on Data, Warehousing and OLAP”, pages 40 to 47.
Hernández-Ros, I. February 4th, 2009. XBRL Infoset 0.3.
XBRL International Inc. Public Working Draft 04
February 2009.
Hernández-Ros, I. and Wallis, H. April 26th, 2010. XBRL
Dimensions 1.0. XBRL International, http://www.xbrl.
org/Specification/XDT-CR3-2006-04-26.rtf.
Kurtev, I. and Berg, K. September 16th, 2003. Model
Driven Architecture based XML Processing.
Proceedings of ACM Symposium on Document
Engineering. September 16th, 2003.
Lucas-Torres Torrillas, F., Navarro Gil, S., García Pérez,
G. and Román Birriales, E. March 18th, 2009.
“Modelos Avanzados de Bases de Datos: Almacenes
de Datos y Bases de Datos XML”. Universidad de
Castilla La Mancha, Escuela Superior de Informática,
Ciudad Real.
Martín Quetglás, G. May, 2006. Curso de XBRL. Pearson
Educación. ISBN 9788483223116.
Mazón, J., Pardillo, J. and Trujillo, J. September 1st, 2007.
Applying Transformations to Model Driven Data
Warehouse. Lecture Notes in Computer Sciencie,
Departament of Software and Computing Systems,
University of Alicante, Spain.
Mazón, J., Trujillo, J., Serrano, M. and Piattini, M.
November 1st, 2005. Applying MDA to the
Development of Data Warehouses. DOLAP’50;
Proceeding of the 8th ACM International workshop on
Data Warehousing and OLAP.
Ogbuje, U. May, 2002. XML, The Model Driven
Architecture and RDF. In XML Europe 2002. May
20st-23rd, 2002, Barcelona, Spain.
Pakorný, J. February 11th, 2002. XML Data warehouse:
Modelling and Querying. Proceedings of the Baltic
Conference, Baltic DB&IS 2002, volume 1, pages:
267-280”. Institute of Cybernetics at Tallin Technical
University.
Pardillo, J. and Trujillo, J. October 13th, 2008. Integrated
Model-Driven Development of Goal-Oriented Data
Warehouse and Data Marts. Lucentia Research
Group, Department of Software and Computing
Systems, University of Alicante. In Lecture Notes in
Computer Science.
Piechocki, M., Felden, C., and Gräning, A. March 25th,
2007. Multidimensional XBRL Reporting. Proceedings
of the Fifteenth European Conference on Information
Systems.
Schmehl, K. and Ochocki, B. November 19th, 2009. Data
Model and Matrix Schemas. http://www.eurofiling.
info. XI European Banking Supervisor XBRL
Workshop, 16th November 2009, Vienna.
Snijders, P.. February 11th, 2005. XBRL and Complex
Data Mapping. Madrid, Conference for Software
Developers.
XBRL International, December, 14th, 2001. Extensible
Business Reporting Language (XBRL) 2.0
Specification. http://www.xbrl.org/tr/2001/xbrl-2001-
12-14.doc.
WEBIST 2011 - 7th International Conference on Web Information Systems and Technologies
164