Content uploaded by Milos Jovanovik
Author content
All content in this area was uploaded by Milos Jovanovik
Content may be subject to copyright.
Content uploaded by Milos Jovanovik
Author content
All content in this area was uploaded by Milos Jovanovik
Content may be subject to copyright.
The 7
th
International Conference for Informatics and Information Technology (CIIT 2010)
A KNOWLEDGEBASE EDITOR USING SEMANTIC WEB
TECHNOLOGIES
Goran Bakraceski, Riste Stojanov, Milos Jovanovik, Dimitar Trajanov
Faculty of Electrical Engineering and Information Technologies
Rugjer Boskovik bb, 1000 Skopje, Macedonia
goran.bakraceski@gmail.com, ristes@feit.ukim.edu.mk,
milos@feit.ukim.edu.mk, dimitar.trajanov@feit.ukim.edu.mk
ABSTRACT
We agree that a semi - structured data model offers the right
balance of rich structure and flexible schema allowing users
to store and manage data as they prefer, making the access as
easy as possible. For that reason we build this system, which
exposes the richness and flexibility of the data model,
offering natural access to the data and hides the specifics of
schema, tuples and database queries. This system uses
semantic web technologies for data representation, providing
form and text based data entry and retrieval. Data is stored in
RDF (Resource Definition Framework) format [1], while the
forms are generated from the OWL (Ontology Web
Language) data definition [2]. The OWL based guiding is
provided for the text entry via suggestions. Users with poor
knowledge in OWL and RDF can use this system for defining
the format of their data, by entering and altering OWL
classes, and manipulate with their instances.
Keywords: silverlight application, semantic web,
knowledgebase editor
INTRODUCTION
Nowadays one of the most interesting topics in modern web
technology is the semantic web and the opportunities that it
has. The computers are handicapped when they need to
understand information. The information is all around us, but
for now, only people can understand it. the semantic way is
introducing a new way of knowledge representation, not only
for the user,bat to the machines too. The main idea is to create
schemas, that will hold the definition of the data used in the
application. But, this definition is not as abstract as the class
definition in the programming languages. The semantic web
formats are designed to be close to the natural language, so
that they can be understandable for regular people, and not
only by the experts. Thus, the people can easily use and
manipulate the data, and define its structure using structure
that is very close to their natural language.
Using this new technology, development of new GUI
(Graphical User Interfaces) can be semantically driven and
users can use and manage data in their preferable way.
However, we have to mention that this technology is quite
new and users doesn't have much knowledge for the data
representation, even it is similar to our way of communication
whit each other. One of the main objectives of this
application, among others, is exactly that, to provide easy and
simple interface for creating new ontology and changing
existing one. For providing the user flexibility in what they
store and retrieve, this application communicates with
knowledge management system (KMS). Objects are
identified by uniform resource identifier (URI), which
guaranty the uniqueness of the resource. Using this URI, user
can reference any resource of its interest, while he builds his
own knowledgebase. In that way, users are exempt from
learning methods for creating ontologies and schemas, simply
because the application provides suitable way for it. On the
other hand, with this application we can link the knowledge,
offering one resource to be significant for the other.
At the beginning of this paper, we will outline few
applications with similar behaviour and goal as ours. They
use the advantages and benefits of the semantic web for
storing data and creating ontology. Then, we will summarize
the primary goals and the motivation for building this
application, and present the the web based guided text editor
and the form based editor. The paper is closed with a section
on related work and some conclusions.
RELATED WORK
The developed web based guided text editor aims to improve
user ability for creating and editing ontology, together with
their instances, using guided natural language form, with
typing natural language sentences. It offers using of auto
complete feature that is guiding the user in the data entry, and
user can store and retrieve his data with no knowledge in
semantic web technology and its formats of knowledge
representation. Similar applications, that are also using the
advantages of the semantic web, are described by Error:
Reference source not foundError: Reference source not found.
In GINOError: Reference source not found the author
describes the application for guided ontology creation and
manipulation using text based interface, which is a Java
application for creating new ontology and editing existing
one. Here, user can see and select any class or object from the
ontology using the application and additionally can edit
selected one. The Haystack Error: Reference source not found
application is much more complex. It uses semantic data for
storing, searching and retrieving data for enabling flexible
user interface and resource linking in a way that user wants.
This application allows users to customize their interface.
Because the User Interface is customizable, this makes users
to be more efficient. Resources from different areas can be
linked together, making this application replacement for
several other applications. User can link its favourite data in
this way:
©2010 Institute of Informatics.
The 7
th
International Conference for Informatics and Information Technology (CIIT 2010)
• Associate e-mail about a certain interview with the
interview article you are writing;
• Link musicians to concerts they played, songs they
performed, and photographs they are in;
• “Caption” photographs of musicians with the song
being performed in the photo;
• Place songs or albums in a calendar according to
release date;
Therefore, the power of using this new way of storing data
and its manipulation, was main motivation for building this
knowledgebase editor. It functionality can be divided in two
sections. In the first section user can create or edit an
ontology using forms. The forms are well defined so, users
with no experience in semantic web can add new classes and
objects, or edit the existing one.
KNOWLEDGEBASE EDITOR USECASE
The user interface is designed using Silverlight 2.0 [6], while
the data manipulation is done using the Jena ontology
framework [8]. Because Jena is developed in Java, the
communication between Jena and the user interface is
achieved by using web services. The two programs
communicate with each other with common objects. These
common objects are all that we need to get information about
ontology classes, properties, instances etc. Currently they are
wrapping the Jena defined objects for the corresponding
instance. We use four types of objects including
CommonClass, CommonProperty, CommonInstance and
CommonStatement. These are sufficient to make the
necessary operations. The architecture and data model will be
described in details in the next section.
At the application startup, all classes and properties from the
ontology are loaded into the Jena model, so the user can
manipulate with them. The interface for data manipulation is
shown in Figure 1.
Figure 1: Class and property manipulation
At the upper - left side of the Figure 1 is shown the part of the
user interface where the user can iterate through all the
classes that are contained in the Jena Model. Here, the user
can edit some of the classes, or, if he wants, he can create a
new one and add it to the ontology. Creating new classes is
easy and simple. Simply he have to type the name for the new
class, choose the super class (inheritance is enabled) and
some additional attributes like comment and label for newly
create class. After this, the name of the new class appears in
the list box that contains all the classes of the ontology. The
second step is to add properties for this class. First of all the
user have to choice one class from the combo box shown in
Figure 1. After that, all properties whose domain is the class
are listed in list box. By selecting one of them, the form
(shown on the right side in Figure 1) is filled with data
specific for the selected property. At this way, data can be
edited and properties can be updated. Inserting new properties
for an existing class is the same. Under the list, filled with all
properties that might have one class, is positioned “New
property” button. After clicking this button, the form for
adding new property is reset and ready for new one. After
filling the form, the new property will be added to the
selected class from combo box that contains all classes.
The next tab is reserved for instance manipulation but the last
is something different. The Instance tab offers the user to look
all the instances that are inside the ontology. By selecting one
of the offered classes, the user can see all the instances of that
class. These instances are listed in list box show in Figure 2.
The next step is selecting one of the instances. By doing that,
the user can see which properties are describe the instance
and he can see how their values. He can make changes and
update the instance simply by clicking on Update button. The
new updated instance is now inserted in the loaded ontology.
Figure 2: Working with instances
Inserting new instances of selected class is also available.
Under the list containing all instances for selected class, is
positioned “New instance” button. By clicking this button, the
form on the right appears, containing all properties subscribed
to selected class. After filling this form, new instance is
created and added to list of all instances, available for
selected class.
Best way to explain the last tab is by using an example.
Let suppose that we have two classes: the Person class and
The 7
th
International Conference for Informatics and Information Technology (CIIT 2010)
the City class. The Person class has property is_from (where
the person lives) and the City class has property name (the
name of the city). The ranges of the is_from property are
instances of type City. We have created two instances of type
Person: Bill who lives in London and Jon whose property
is_from is not set.
Here comes the using of last tab. The user starts typing and he
enter Bill. After he hit space in the auto complete box is
shown is_from. The interface suggested the user that next he
can enter the word is_from. After inserting is_from and
hitting another space, the user interface knows that Bill is
from “London”, and in auto complete box is shown “London”
(left side of Figure 3).
Figure 3: Manipulation with ontology using free
text input
On the other hand if the user starts typing John and choose
is_from from auto complete, after hitting another space
because there is no information where John lives, in the auto
complete box are shown all available cities (in our case
London and Barcelona) and the user can choose one of them.
This example is shown on Figure 3. By doing this, the user is
guided by auto complete feature, which intelligently suggest
and offer next words that might be added in process of linking
and connecting data contained in the ontology.
ARCHITECTURE AND TECHNICAL DESIGN
From an architectural point of view, this application has
two main parts and a few subparts. First part is the knowledge
management system (KMS), that makes persistence and
manipulation with the ontology. Behind this knowledge
management system is the Jena framework that is responsible
for handling request that makes changes to the ontology. In
addition, Jena takes care of making some reasoning based on
the ontology, and due to this operation, new links and
connections between the resources are made. Because this
operation of reasoning is very expensive, in terms of time and
resources, it is not done very often. The other thing that is
more significant for our application are the web services that
this KMS provides. Because our application is Silverlight
application and KMS and Jena were Java applications, the
need of web services was inevitable. Therefore, this KMS
offers a set of few operations, significant for ontology and
data manipulation.
Figure 4: Application architecture and technical
design
Because Microsoft Silverlight 2.0 [6] is technology which
communicate with other applications only via Silverlight –
enabled web services, a translation was necessary. Therefore,
the data that were provided by KMS first were received by
ASP.NET application [7], and then forwarded to the
Silverlight application via Silverlight – enabled web services.
Visually this architecture is represented on Figure 4. After the
data is received, the application displays properly and
performs operations over them. Depending on this operation,
corresponding forms are filled, or assembled, and data is
represented.
Additionally, because operations for retrieving classes and
properties from ontology sometimes may take a while,
caching of data is performed. At the beginning of the
application, before the user interface is displayed, data for
classes and its properties are imported in the application and
cached. Therefore, operations where manipulations with
classes and properties are needed, they are read from cache
rather than making new call to web services. Updating data is
made on both sides, on cached data and on data represented
by the ontology.
ADVANTAGES AND DISADVANTAGES
We agree that every application has advantages and
disadvantages. As first advantage of this application can be
mention good way of caching data. That way the application
runs faster, transparently executing operations for data
manipulation. Next is the auto-complete feature that guides
the user while he is writing natural sentences and tries to
connect the resources. In addition, we have to mention that
the forms for creating and editing classes, properties and
instances are pretty simple and clear, so, the user can easily
manipulate with the ontology. Another advantage is the
transparent generation of the forms for entry and
manipulation with the ontology resources. Thus we have a
suitable form for every possible type of resource.
As a disadvantage of this application can be mentioned
the inequality of the sequence by which the fields are created
The 7
th
International Conference for Informatics and Information Technology (CIIT 2010)
when ontology instance is created. The order of the fields
depends on how much properties have given instance, and its
type. Listing all properties when creating an instance is also
disadvantage of the user interface. There is a class that can
contain several properties and filling all of them can be hard
and painful. This can be resolved by using views and
collections in some of the next improvements of the
application.
CONCLUSION
With the work presented in this paper we achieve flexible
system for data entry, that utilizes the benefits of the semantic
web technologies. Due to the SOA [5] oriented design of this
system, it is easier to maintain, and its parts can be reused in
other systems, providing eater flexible data presentation and
entry, or efficient manipulation with semantic data.
FUTURE WORK
Because the possibilities that offer the semantic web are huge,
research in this area does not stop here. Our aim is to improve
interface interactivity with the user, facilitating the way of
entering date in process of building new ontology or editing
existing one. Making heterogeneous user interface, which
would be combination of forms and free typing, would make
data entry very easy for the end user. Other good feature
would be graphically display of the entire ontology, so the
user can see the relationship of the data and its connection.
We hope that this will be done in near future, and new editor
with better performance and reasoning will be made.
REFERENCES
[1] RDF, Resource Description Framework, http://www.w3.org/RDF
[2] OWL, Web Ontology Language, http://www.w3.org/TR/owl-
features/
[3] Abraham Bernstein and Esther Kaufmann: GINO - A Guided
Input Natural Language Ontology Editor, University of Zurich,
Dynamic and Distributed Information Systems, Switzerland, 2007
[4] David R. Karger Karun Bakshi David Huynh Dennis Quan
Vineet Sinha. Haystack: A Customizable General-Purpose
Information Management Tool for End Users of
Semistructured Data,MIT Laboratory for Computer Science,
2002
[5] Krafzig, D. and Banke, K. and Slama, D., Enterprise SOA:
Service-Oriented Architecture Best Practices (The Coad Series),
Prentice Hall PTR Upper Saddle River, NJ, USA, 2004
[6] Silverlight 2.0, http://silverlight.net/
[7] ASP.Net, www.asp.net
[8] Jena – A Semantic Web Framework for Java,
http://www.jena.sourceforge.net
