SEMANTIC WEB TECHNOLOGIES IN EEG/ERP DOMAIN
Software Solution
Roman Mouček, Petr Ježek and Václav Papež
Department of Computer Science and Engineering, University of West Bohemia
Univerzitní 8, 306 14 Pilsen, Czech Republic
Keywords: Semantic web, RDF, OWL, Relational database, Automated transformation, Electroencephalography,
Event-related potentials, EEG/ERP portal, Jena, D2RQ, OWL API.
Abstract: The paper shortly describes EEG/ERP domain and the system for storage and management of data and
metadata from EEG/ERP experiments. The system has been developed using common technologies and
data structures (object oriented code, relational database). A registration of the system as a recognized data
source requires representation of its data and metadata by semantic web resources. A short overview of
differences between common data structures and semantic web representations is given and possible
approaches to automated transformation of data and metadata from common data structures are mentioned.
The existing semantic web frameworks and tools, widely tested for our purposes, are listed. Finally, the
integration of selected tools and the software solution for automated transformation from the relational
database to a semantic web representation is described.
1 INTRODUCTION
Our research group at Department of Computer
Science and Engineering, University of West
Bohemia in cooperation with other partner
institutions (Czech Technical University in Prague,
University Hospital in Pilsen, Škoda Auto, Inc....)
specializes in the research of attention, especially
attention of drivers, children with developmental
coordination disorders and seriously injured people.
We widely use the methods of
electroencephalography (EEG) and methods of
event-related potentials (ERP).
We are members of International
Neuroinformatics Coordinating Facility (INCF)
(INCF, 2010). Among other things this organization
supports collaboration of national nodes for
neuroinformatics and standardization of data and
metadata structures in the domain of
neuroinformatics.
We have developed a software tool for a long
term storage and management of EEG/ERP
experiments. This tool is based on a relational
database and modern Java technologies.
A registration of this data source within the
neuroinformatics community requires providing data
and metadata in a form of ontology. This led us to
the idea of using semantic web technologies.
However, our goal is to provide this ontology
automatically from the existing structures, not to
build a separate ontological model.
The aim of this paper is to provide readers with
our approach in usage of semantic web resources,
techniques, frameworks and tools in the EEG/ERP
domain. Finally, a presentation of our software
solution is given.
The paper is organized as follows. At first we
briefly introduce EEG/ERP domain. Then the
system for storage and management of EEG/ERP
experiments is shortly described. Possibilities for
transformation of data and metadata from common
data structures are mentioned and some known
approaches to a mapping of knowledge from
common data structures (relational database, object
oriented code) to semantic web languages are
summarized. Existing mapping tools, which were
widely tested, are listed and the final software
solution for a transformation of relational data to
a semantic web representation is described.
618
Mou
ˇ
cek R., Ježek P. and Papež V..
SEMANTIC WEB TECHNOLOGIES IN EEG/ERP DOMAIN - Software Solution.
DOI: 10.5220/0003291406180621
In Proceedings of the International Conference on Health Informatics (HEALTHINF-2011), pages 618-621
ISBN: 978-989-8425-34-8
Copyright
c
2011 SCITEPRESS (Science and Technology Publications, Lda.)
2 EEG/ERP DOMAIN
EEG and ERP methods are widely used in research
of brain processes. EEG and ERP experiments take
usually long time and produce extensive data.
Experiments made in our laboratory were initially
recorded and managed using a common folder
structure (commercial software tools are used during
recordings). Because an appropriate software tool
for a long-term storage and management of data
obtained during EEG/ERP experiments did not exist,
we have developed a custom software solution. This
software tool, EEG/ERP portal, is now used in our
department and it will be provided to
neuroinformatics community after more extensive
testing.
The basic ideas and features of the system are
described in (Ježek, 2010). From that time the
software incorporated a number of changes and
improvements. Not only EEG/ERP experiments but
also data and metadata structures in various formats,
scenario templates of EEG/ERP experiments, groups
of people involved in experiments, etc. are now
covered by the features of EEG/ERP portal.
However, these changes and improvements are not
important for the aim of this paper.
3 EEG/ERP PORTAL
EEG/ERP portal enables community researchers to
store, update and download data and metadata from
EEG/ERP experiments. The system is developed as
a standalone product; the database access is
available through a web interface.
The system essentially offers the following set of
features (the set of accessible features depends on
a specific user role):
User authentication
Storage, update, and download of EEG/ERP
data and metadata
Storage, update and download of EEG/ERP
experimental design (experimental scenarios)
Storage, update and download of data related to
testing subjects
Sharing of knowledge and working in groups
The three layer architecture (MVC pattern) was used
to build the EEG/ERP portal. It consists of the
persistent layer (relational database), the application
layer (object oriented code, object relational
mapping from persistence layer) and the
presentation layer (JSP).
4 MAPPING TO SEMANTIC WEB
LANGUAGES
Scientific papers often describe a domain using
a semantic web language and consider this kind of
domain modelling as a crucial point of software
solution. However, real software applications use up
the underlying data structures such as relational
databases or object classes. Then a mapping from
common data structures to a semantic web language
is necessary.
Concerning the architectural layers of EEG/ERP
portal there is a question which layer is more
feasible for mapping into a semantic web language
(RDF, OWL). We considered two possibilities:
Mapping from the persistence layer
(relational database)
Mapping from the application layer (object
oriented code)
The mapping from the relational database is
straightforward while the mapping from the
application layer to a semantic web language
requires the precedent object relational mapping
(provided by Hibernate framework in our case).
Since our goal is to generate an ontology, we will
consider OWL as the target semantic web language.
However, it is important to emphasize that OWL
is neither a database framework nor a programming
language. There are fundamental differences in
richness of semantics between OWL (Description
Logic based system) and relational database or
object oriented systems. If some information is not
present in a database, it is considered to be false
(closed-world assumption). By contrast, if some
information is not present in an OWL document, it
may be missing and possibly true (open-world
assumption). On the other hand, there are several
approaches how to bridge at least some of the
semantic gaps. A more detailed overview you can
find in (Mouček, 2010).
5 SEMANTIC WEB
FRAMEWORKS AND TOOLS
There are many frameworks and software tools,
which are considered to generate RDF or OWL
output from a relational database or an object
oriented code. Some of these frameworks and tools
exist only as initial proposals or prototypes
described in scientific papers, while some of them
have been really implemented. The following list
SEMANTIC WEB TECHNOLOGIES IN EEG/ERP DOMAIN - Software Solution
619
includes existing frameworks and software tools,
which were considered to be used for our purposes.
More detailed information about these frameworks
and tools and our experience with them was already
given in (Mouček, 2010):
Jena
D2RQ
Virtuoso
SquirrelRDF
MetaMorphoses
JRDF
SPASQL
Sommer
JenaBean
Java2OWL-S
OWL API
6 MAPPING FROM
RELATIONAL DATABASE
As we already mentioned our goal is to register our
system as the recognized data source and to provide
the system ontology. It means that we need to
perform only one-sided mapping from a relational
database (object oriented code) to OWL; then we
need to use only a subset of semantic richness of
RDF(S) and OWL.
We decided to try out two parallel approaches.
The first approach includes the transformation of
relational database into ontology using D2RQ tool,
Jena and OWL API. The second approach includes
the transformation from object oriented code to
OWL using JenaBean tool.
We focus on the first approach; the second
approach to transformation process is out of scope of
this paper.
6.1 Integration of Selected Tools
We selected three tools, which were finally
integrated to provide the domain ontology from the
relational database of EEG/ERP portal. Jena, the
most widespread tool, was selected for inner
representation of RDF graph. Sesame framework
was rejected because of non-standard approach to
stored data. Moreover, the next selected tool, D2RQ,
cooperates with Jena.
EEG/ERP raw data are stored in the database as
binary files. D2RQ as the second selected tool had
difficulties with BLOB and CLOB data types. The
transformation process generated program
exceptions in this case. We solved it by modification
of D2RQ source code. BLOB data types (raw data)
are processed but they are not included into RDF
graph. CLOB data types (e.g. XML scenarios) are
also processed but they are included into RDF graph.
OWL API was used for storing RDF in various
formats.
The integration of selected tools and the
transformation process can be described as follows.
D2RQ connects to the relational database,
determines its structure and loads this structure to
the virtual RDF graph, which is derived from Jena
graph. Jena transforms this graph into RDF/XML
format. OWL API parses RDF/XML format and
converts it to a format required by a user. Then it is,
for example, possible to transform relational data to
ontology provided in OWL/XML format. It is clear
that this ontology is semantically as rich as the
corresponding relational database.
6.2 Software Description
The final software tool is divided into two parts –
the application library and the graphical user
interface. The application library defines a method
for transformation of the relational database to RDF
graph and a method, which stores this graph into
a resulting file according to a selected format. The
UML diagram of both parts is shown in Figure 1.
Figure 1: Software tool for transformation of relational
database to semantic web representation - UML diagram.
The graphical user interface provides a simple
and transparent way to control the application
(Figure 2).
HEALTHINF 2011 - International Conference on Health Informatics
620
6.3 Formats of Output Files
It is possible to generate the system ontology in the
following formats: RDF/XML, OWL/XML, Turtle,
Manchester OWL syntax, KRSS2, Latex, and OWL
functional syntax. The following example shows
a record of one person from table PERSON in Turtle
syntax.
Figure 2: Software tool for transformation of relational
database to semantic web representation - graphical user
Interface.
###http://eeg_domain/database/EEGERP.PE
RSON/22
<http://eeg_domain/database/EEGERP.PERS
ON/22>
rdf:type
<http://localhost/vocab/EEGERP.PER
SON> ,
owl:NamedIndividual ;
rdfs:label "EEGERP.PERSON #22" ;
<http://localhost/vocab/PERSON_PERSON_I
D> "22"^^xsd:decimal ;
<http://localhost/vocab/PERSON_SURNAME>
"Papez" ;
<http://localhost/vocab/PERSON_GIVENNAM
E>"Vaclav" ;
<http://localhost/vocab/PERSON_AUTHORIT
Y>"ROLE_ADMIN" ;
<http://localhost/vocab/PERSON_PASSWORD
>"bf2bc2545a4a5f5683d9ef3ed0d977e0";
<http://localhost/vocab/PERSON_USERNAME
>"vpapez" ;
<http://localhost/vocab/PERSON_EMAIL>"v
papez@kiv.zcu.cz";
<http://localhost/vocab/PERSON_GENDER>"
M" .
Currently the system ontology is automatically
generated from 23 tables, which are created and
maintained in the relational database of the
EEG/ERP portal.
7 CONCLUSIONS
The presented paper shortly describes the system for
storage and management of data and metadata from
EEG/ERP experiments. This EEG/ERP portal has
been developed using common technologies and
data structures (object oriented code, relational
database). A registration of the portal as
the recognized data source requires representation of
its data and metadata using semantic web resources.
A short description of differences between
common data structures (relational databases and
object oriented code) and semantic web
representations (OWL) was given. The existing
semantic web frameworks and tools, widely tested
for our purposes, were listed. The integration of
selected tools and the final software application that
allows automated transformation from the relational
database to a final semantic web representation were
described.
In the near future we consider finishing
a transformation from an object oriented code to
a semantic web representation. We will also
investigate possibilities of both side mapping
between common structures and widely used
semantic web representations.
ACKNOWLEDGEMENTS
This work was supported by Grant no. 2C06009
Cot-Sewing.
REFERENCES
International Neuroinformatics Coordinating Facility
(INCF). (2010). Retrieved October 22, 2010, from
http://www.incf.org/.
Ježek, P., Mouček, R., 2010. Database of EEG/ERP
experiments. In HEALTHINF 2010 - Proceedings of
the Third International Conference on Health
Informatics. Valencia, Spain.
Mouček, R., Ježek, P., 2010. System for storage and
management of EEG/ERP experiments – generation of
ontology. Proceedings of the 12th International
Conference on Enterprise Information System.
Madeira, Portugal.
SEMANTIC WEB TECHNOLOGIES IN EEG/ERP DOMAIN - Software Solution
621
