BIOMETRIC ACREDITATION ENTITIES
An Approach for Web Acreditation Services
B. Ruiz, L. Puente, D. Carrero
Universidad Carlos III de Madrid, Avda de la Universidad, 30, Leganés, Madrid, Spain
M. J. Poza
Universidad Francisco de Vitoria. Ctra. Pozuelo-Majadahonda Km. 1.800. Pozuelo de Alarcón, Madrid, Spain
Keywords: Biometrics, Authentication Devices, Emerging Technologies, Data Integration, Semantic Web Services,
Ontologies.
Abstract: Identity verification is nowadays a crucial task for security applications. In the near future organizations
dedicated to store individual biometric information will emerge in order to determine individual identity.
Biometric authentication is currently information intensive. The volume and diversity of new data sources
challenge current database technologies. Biometric identity heterogeneity arises when different data sources
interoperate. New promising application fields such as the Semantic Web and Semantic Web Services can
leverage the potential of biometric identity, even though heterogeneity continues rising. Semantic Web
Services provide a platform to integrate the lattice of biometric identity data widely distributed both across
the Internet and within individual organizations. In this paper, we present a framework for solving biometric
identity heterogeneity based on Semantic Web Services. We use a multimodal fusion recognition scenario
as a test-bed for evaluation.
1 INTRODUCTION
Nowadays, the most popular method to gain access
to restricted physical areas is to show to the security
agent the card that identifies the owner as a
privileged person with enough rights go into the
area. Personal cards usually have the owner’s
photography which is a biometrical sample of his
face. On the other side to gain access to restricted
virtual areas, the common use is to apply techniques
such as PINs, passwords, digital signatures, etc.
Losing the personal card, forgetting the password or
PIN or whatever data bearing in mind that performs
personal identification is not only an inconvenient
but also an outrage against the restricted area
security.
Why not to putting these two methods together in
order to get the advantages provided by each one?
Biometrics promises to offer a new alternative,
portable, easy to use, free of memory, loss or theft
problems (Puente et al, 2008). This technology
allows the use of personal traits for individual
identification, as human security agents do, but it
also allows full automatic computer process.
In biometrics, knowledge area authentication
usually means confirming that someone is who
he/she says to be, basing it on his or her
distinguishing traits. It is assumed that these traits
are extracted from personal features. These features
should have certain characteristics that permit
computer processing, such as being measurable,
repeatable by the owner, unrepeatable by others.
Since biometric identity technologies try to solve
security problems dealing with private data, the
challenge for the research community is to attain
integrated solutions that address the entire problems
from sensors and data acquisition to biometric data
analysis preserving personal information.
Currently, in order to increase the accuracy of
the biometric authentication result process, a
technique called “Multimodal Fusion” is being used.
It refers to simultaneous processes of various
samples of different kind of traits (Kittler et al,
1998) (Jain et al, 2001). As a result of this, biometric
information has grown exponentially and algorithms
216
Ruiz B., Puente L., Carrero D. and J. Poza M. (2008).
BIOMETRIC ACREDITATION ENTITIES - An Approach for Web Acreditation Services.
In Proceedings of the International Conference on Signal Processing and Multimedia Applications, pages 216-220
DOI: 10.5220/0001939902160220
Copyright
c
 SciTePress

for feature extraction, matching score or decision
levels handle a tremendous amount of data.
Furthermore, the recent years have provided an
amount of duplicated efforts in building test
databases such as face recognition databases (e.g.
FERET, PIE or BANCA) (Bailly-Baillière et al,
2003) as well as a lack of uniform standards and
granted open access to these databases, as discussed
in (Ming et al, 2007).
Hence, the most critical need in biometric
identity recognition is arguably to overcome
semantic heterogeneity i.e. to identify elements in
the different databases that represent the same or
related biometric identities and to solve the
differences in database structures or schemas, among
the related elements. Such data integration is
technically difficult for several reasons. First, the
technologies which different databases are based on
may differ and may not interoperate smoothly.
Standards for cross-database communication allow
the databases (and their users) to exchange
information. Secondly, the precise naming
conventions for many scientific concepts in fast
developing fields such as biometrics are often
inconsistent, and so mappings are required between
different vocabularies.
Figure 1: Model generation.
Therefore, in this paper we talk about the
feasibility of taking advantages from a biometrics’
technology framework preserving individual rights.
The remainder of this paper is organized as
follows. Section 2 compiles a brief list of terms used
along this paper. Section 3 introduces the current
environment for the biometric works. Section 4
proposes a new environment for authentication
purposes. Section 5 describes an experimentation
environment. Section 6 compiles authors’
conclusions.
2 USED TERMS
Before continuing talking about biometrics it is
necessary to fix some terms that we are going to use
along this paper.
A trait is defined as any physical, motor or
psychomotor human characteristic capable of being
used in biometric identification.
A user is any person to be recognized by the
system, and whose traits are somehow stored in the
database.
A donor is any person (user or not) whose trait is
captured, voluntary or involuntary, by a sensor of
the system.
A sample is defined in (Mansfield et al, 2002) as
a biometric measure presented by the donor which
eventually results in an image or signal.
Feature refers to a mathematical/measurable
characteristic of the acquired sample. Sometimes it
is the sample itself, other times it is the result of a
more or less complicated mathematical process, in
that case the process often captures a set of different
features called feature vector. This process is known
as “feature extraction”.
3 CURRENT ENVIRONMENT
A typical biometric system presents a well defined
structure (Mansfield et al, 2002) that includes two
phases: enrolment and testing.
Enrolment faces the creation of a type of model
representing the user in a univocal way. It starts
storing a sample of one of the biometrical traits (data
acquisition) from the sensor output data, the
following process (feature extraction) tries to extract
from this sample information that univocally
characterizes the individual, avoiding to include the
variable components of the trait and obtaining a set
of feature vectors. At the end a (mathematical or
not) model is obtained. Sometimes this
transformation is trivial and the result model is the
sample itself. Other times it is no so simple and at
the end we obtain a mathematical expression with
the set of its coefficients (see Figure 1).
On the other hand testing starts with the same
steps of data acquisition and feature extraction
obtaining a set of feature vectors (see Figure 2). But
at the end it matches the vector feature set with the
stored model of the individual. Model matching
establishes a metric system in which a distance
between the model and the sample is defined, the
longer this distance the more probable is that the
donor is not the model’s user. The distance is
BIOMETRIC ACREDITATION ENTITIES - An Approach for Web Acreditation Services
217

compared with a threshold (that leads to the
decision) deciding if the donor is the user
(acceptance) or not (rejection).
Figure 2: The testing process.
For the two phases (enrolment and testing)
acquisition implies that one or more sensors acquire
one or more samples of certain donor’s biometric
traits presented to the biometric systems (e.g.
fingerprint, face, iris image) (Puente et al, 2008).
Two different tasks should be executed inside
this environment. In the first one, called
authentication, a donor declares to be a certain user
(claimed user) and the system determines if it is true
or not by checking if the donor’s biometrical sample
matches or not the user model.
In the second one, called identification, the donor
does not inform about his/her identity, usually
because he/she is an involuntary donor. Then it is
necessary to look for the best accuracy models from
all of the stored ones in the database and it will
determine the probability of being any of the model
users.
Nowadays user models are stored in a very local
database. Multimodal fusion increases the
complexity forcing to store no less than a model for
each modality and for each user to be authorized.
Furthermore the user has to place in every biometric
database of the restricted areas he/she needs to
access, no less than a biometric sample for each
modality. It is not only an inconvenient but a
problem of information protection, because this
redundancy increases the probability that the
biometric data stays in unauthorized hands, and
makes practically impossible to remove this
information from all databases.
Experimentation framework shows this problem
in a more critical way. Currently, some entities have
compiled databases of biometric samples. These
samples were obtained from anonymous donors,
who often give their authorization only for biometric
experimentation. This database has been shared by
researches and the usual support is a CD-ROM. In
this environment it is impossible to control the use
made with this information, or to remove certain
samples from it.
4 EXPERIMENTAL RESULTS
Nowadays, several classification system (Dessimoz
et al, 2006) and fusion techniques (Jain et al, 2005)
exist in order to verify persons’ identity.
The results obtained in identity verification using
fusion of biometric data at score level from iris,
signature and voice are shown in
TABLE . The error
rate of unimodal biometrics systems are 12.4%,
5.73% and 25% respectively.
These results show the capability of recent fusion
techniques to reduce the error rate in identity
verification tasks. In this way, multimodal
biometrics becomes one of the main tools in BAE.
5 PROPOSED ENVIRONMENT
FOR AUTHENTICATION
SERVICES
We propose not to share de data but the services.
A global solution will be based on the creation of
specialized organizations offering authentication
services. Of course, this Biometric Accreditation
Entities (BAE) will obviously base their services on
previously acquired biometric data. Then BAE could
be created as specialized organizations dedicated to
collect and store individual biometrical information
and to offer identity accreditation services.
Throughout the Web BAEs will supply identity
accreditation of their registered users just for user
authorized organization.
The major limitation for BAE being useful is
that all over the world we can find very different
Table 1: Rank of Methods Error Rate.
Fusion Method
Error
Rate
3-3-1 N. Network with simple normalized data
1.58%
3-3-1 N.Network with normalization and
sigmoid transformation
1.65%
Weighted Product with simple normalized data
and Dynamic Score Selection
1.66%
Weighted Product with simple normalized data
1.67%
SVM with normalization and sigmoid
transformation
1.68%
SIGMAP 2008 - International Conference on Signal Processing and Multimedia Applications
218

techniques to capture biometric traits and therefore
getting heterogeneous data formats.
The first option in order to face this problem is
to create standards that normalize biometric data to
be exchanged with BAE. In opposition to that, our
proposal includes, as well as the BAE concept, the
use of “Semantic Web Services” (SWS).
In this environment, BAE will provide a platform
that allows data matching of acquired biometric
samples, encapsulated in a semantic description
bubble, against individual biometric models stored
inside the BAE.
In addition to the above service, BAEs will also
provide catalogues of the stored data. Not only does
it allow determining where the models for a given
user are located but also where the most accurate
one for data acquisition process is.
Continuing in the development of this concept
our future work will focus the creation of the
semantic context by the definition of the ontology
oriented to the purpose of making it possible.
The main application scenarios for authentication
services are listed below. By one hand, military and
defense scenarios can be described:
• Security in airports: BAE verification is a
complementary way to improve traditional
verification systems like passport, ID card and
driving license.
• Frontiers control: BAE can offer support to the
surveillance carried out by the security agents in
order to identify and classify travelers in the
entrance points to the countries.
• Access control to restricted areas: BAE offers
second user verification behind the control made by
security staff.
By the other hand, civil applications can be
described too.
• Credit card payment: In order to complete an
electronic transaction, BAE user verification is
needed next to the user’s financial data and
signature.
• Documents accreditation: BAE may validate
authorship of electronic documents by including a
verification certificate about the author's identity.
This system is similar to digital signature
procedures.
• Control over employees: BAE may assist into
the recognition tasks over employees. In the same
way, BAE may improve the time control tasks over
employees. Biometrics rules out in an effective
manner some situations in which a person uses an
identification card to prove other person's presence.
• Nurseries: BAE may verify parents’ identity at
the moment they pick their children.
In order to check the validity of donor’s
biometric data, biometric data acquisition have to be
supervised by humans.
6 PROPOSED ENVIRONMENT
FOR EXPERIMENTATION
Previously, we have mentioned that researchers on
biometrics suffer the same problem related to the
need of having access to biometrical databases.
In this context our proposal aim is similar to the
one above for BAE: “not to share the data but the
services”.
In this case we promote the creation of a
platform called “Biometrical Extended Experiment
Platform” (BEEP), which is a virtual space where
researchers can test their algorithm against a large
multimodal database.
BEEP offers not only the database but the
possibility of a true comparison with the published
results of other colleges obtained in a similar,
controlled and normalized environment.
It is BEEP’s responsibility to ensure that only
global results are at the end of a BEEP process, and
that no individual information can be obtained.
As a continuation of this, our future work will
define the algebraic context of BEEP in order to
make it available for researchers, and to create its
inside database.
7 CONCLUSIONS AND RELATED
WORK
The creation of Biometric Accreditation Entities will
be an alternative in the near future to the current
digital certification organisms.
In this environment the heterogeneity of sample
capture and data process should not become a barrier
for the use of this identification technology.
In order to do that, Ontology and Semantic Web
Services show their capabilities to offer a solution to
the growing problem of the heterogeneous data.
On the other hand biometrical researching could
be normalized by BEEP, offering a platform for
making results’ comparison.
Finally, our future work will focus on creating a
completely adapted ontology and defining a standard
for required services on SWS as well as determining
an algebraic context for the biometrical
experimentation process.
BIOMETRIC ACREDITATION ENTITIES - An Approach for Web Acreditation Services
219

ACKNOWLEDGEMENTS
This work is founded by the Ministry of Science and
Technology of Spain under the PIBES project of the
Spanish Committee of Education & Science
(TEC2006-12365-C02-01).
REFERENCES
Bailly-Baillière, E., Bengio, S. et al. 2003. The BANCA
Database and Evaluation Protocol, in Springer LNCS-
2688, 4th Int. Conf. Audio- and Video-Based
Biometric Person Authentication, AVBPA’03. 2003,
Springer-Verlag.
Berners-Lee, T., Hendler, J. and Lassila, O. 2001. The
Semantic Web. Scientific American, May 2001, pp.
34-43.
Borst, W.N. 1997. Construction of Engineering Ontologies
for Knowledge Sharing and Reuse. PhD Thesis.
University of Twente. Enschede, The Netherlands.
Dessimoz, D., Champod, C., Richiardi, J. and Drygajlo, A.
2006. MBIOD. Multimodal Biometrics for Identity
Documentation. State-of-the-art. Research Report.
University of Lausanne. Available at:
http://www.europeanbiometrics.info/images/resources/
90_264_file.pdf.
Fensel, D. and Bussler, C. 2002. The Web Service
Modeling Framework WSMF. Electronic Commerce
Research and Applications, 1(2).
Gandon, F. and Sadeh, N. 2004. Semantic Web
Technologies to Reconcile Privacy and Context
Awareness. Web Semantics Journal, 1(3), 2004.
Gibbins, N., Harris, S. and Shadbolt, N. 2003. Agent-
based Semantic Web Services. In Proc. of the 12th Int.
World Wide Web Conference, May 2003.
Gómez, J. M., Rico-Almodóvar, M., García-Sánchez, F.,
Martínez-Bejar, R. and Bussler, C., 2004. GODO:
Goal-driven Orchestration for Semantic Web Services.
WSMO Implementation Workshop, September 2004.
Gruber, T. R. 1993. A translation approach to portable
ontology specifications. Knowledge Acquisition vol.
5:199-220.
Hendler, J. 2001. Agents and the Semantic Web. IEEE
Intelligent Systems, 16(2): 30-37, March/April 2001.
Jain, A., Nandakumar, K. and Ross, A. 2005. Score
Normalization in Multibiometrics Systems. Pattern
Recognition, vol. 38, pp. 2270-2285, Jan. 2005.
Jain, K., Bolle, R. et al.1999. Personal Identification in
Networked Society. Kulwer Academic. 1999.
Jain, R. and Quian, J. 2001. Information Fusion in
Biometrics. Proc. 3rd International Conference on
Audio and Video Based Person Authentication
(AVBPA) pp. 354-391, Sweden, 2001.
Kittler, J. Hatef, R. and Matas, J. G. 1998. On Combining
Classifiers. IEEE Transactions on PAMI, vol. 12
(1998). Pp. 226-339.
Mansfield, J. and Wayman, J.L. 2002. Best Practices in
Testing and Reporting Performance of Biometric
Devices. National Physics Lab for Mathematics and
Scientific Computing. 2002.
Ming, A. and Ma, H. 2007. An Algorithm Tested for the
Biometrics Grid. Proceedings of the Second
International Conference in Grid and Pervasive
Computing (GPC07). Paris, France. 2007.
Puente, L., Poza, M.J., Ruiz, B. and Gómez, J.M. 2008.
Biometric Authentication Devices and Semantic Web
Services. An Approach for Multi Modal Fusion
Framework , BIODEVICES 2008.
Web Ontology Working Group, 2004. OWL Web
Ontology Language Guide.
OWL-S W3C Submission, 2004. OWL Web Ontology
Language for Services. Available at:
http://www.w3.org/Submission/2004/07/
SWSF W3C Submission, 2005. Semantic Web Service
Framework. Available at:
http://www.w3.org/Submission/2005/07/
WSMO W3C Submission, 2005. Web Service Modeling
Ontology. Available at:
http://www.w3.org/Submission/2005/06/
WSDL-S W3C Submission, 2005. Web Service
Semantics. Available at:
http://www.w3.org/Submission/2005/10/
SIGMAP 2008 - International Conference on Signal Processing and Multimedia Applications
220