Implementation of Ontology in Intelligent E-learning System

Development based on Semantic Web

S. RAM KANNAN, PRASHANTH ANAND, R SABARISH, SARAVANAN
Computer Science and Engineering, IV year
CEG, Anna University
Chennai-100.

Abstract : Semantic Web for the provision of distributed

It is visualized that Ontologies and Semantic Web information with well-defined meaning,
technologies will influence the next generation of understandable for different parties based on
E-learning environment. Ontology provides flexible Ontology combines the of context, content and
educational platform architecture for E-learning structure of the learning materials, providing flexible
environment. The success of the Semantic Web access to these learning materials. Development of
depends on the procreation of ontologies. In Ontology-based E-Learning support system by using
general, Ontology based Learning supports the a Semantic Web allows learners to build dynamic
construction of ontologies by an ontology engineer. learning paths and how the Semantic Web resource
It takes over the ability to act in a genuinely description formats can be utilized for automatic
intelligent manner by assessing the learners initially generation of hypertext structures from distributed
and providing personalized suggestions to the metadata.
learners indicating their strengths and weaknesses. This project focuses on the development of
This ontology learning framework proceeds through ontology-based E-learning support system which
ontology importing, extracting, pruning, allows learners to build dynamic learning paths. We
refinement, and evaluating giving the ontology also present an approach for developing a Semantic
engineer a wealth of coordinated tools for ontology Web-based E-learning system, which focus on Web
modeling This tool has been developed with the Ontology language RDF and OWL. Additionally an
view of providing software based training to automated learning service system has been
potential engineers and educating them. We developed. Our e-learning support system project
introduce our ontology model and propose an aims at creating, integrating and interfacing of
effective method for enhancing learning effect of multiple ontologies on different layers, such that all
students through construction of subject ontology. are grouped together under a single ontology called
This paper focuses on the development of domain ontology. This tool has been developed with
ontology-based E-learning support system which the view of providing industry-oriented training to
allows learners to build dynamic learning paths. We potential engineers and educating them. It gives
also present an approach for developing a Semantic flexible educational platform architecture for
Web-based intelligent E-learning support system, E-learning system.
which focus on the Web Ontology Language RDF
and OWL. Additionally an automated and an The proposed work includes:
intelligent e-learning service system have been 🢡 Structured E-Learning Websites – with
developed. relationship.
🢡 Effective Search engines with exact result.
Keywords – E-learning, Semantic Web, RDF, OWL, (along with dynamic keyword prediction)
Ontology. 🢡 Includes Semantic Web that makes
searching user-friendly (Directly tied to the
specificities of the web environment).
🢡 To build Dynamic Learning Paths through
understanding curriculum.
I. INTRODUCTION 🢡 Tests for the levels or courses completed and
Score cards mailed to the test takers.
E-LEARNING, refers to online learning or distance 🢡 Best material for people preparing for the
learning, allows users to access electronic learning placements since it takes them across a tour
contents delivered over internet or intranet. It is to the necessary subjects.
concerned with the development of efficient
computer-aided education system. Our aim is to use
 🢡 Making use of OWL(“Semantic Web Repository is the database of the entire system. It is
Vision”) increases the machine the place where the RDF files, OWL files are present.
interpretability. It contains the entire details of data present. It is the
🢡 YouTube videos with online or offline heart of this ontology system where there is complete
streaming and External Website Contents. presence of the entire data. Any information required
E-Learning content are directly tied to the will be fetched only from this repository. It also
specificities of the web environment as shown in the includes the student database which contains all the
architecture diagram. All the Semantic Web required details of the students who enter into this
technologies are used in this paper. system. Any student’s complete details will be
present in this student repository.
II. SYSTEM ARCHITECTURE
ADMINISTRATOR:-
The architecture of the developed e-learning system
consists of four important parts namely : The administrator is the sole person who controls the
entire functioning of the system. The administrator
● Learning environment appoints people to take the sub functions namely
● Semantic web evaluator, instructor, advisor where they perform
● Repository roles like evaluating the performance of the students
● administrator based on the tests they take, instructor instructing the
students with the functionality of the system.

LEARNING ENVIRONMENT:-
The learning environment consists of all the events
that prevail in the system for students. The
registration work consists of the process were the
students who wish to take up their learning have to
first register themselves with all the required details
and information. This will gain them an entry to view
the system that contains the course details. The
course details consist of all the necessary details that
the student wishes to take. It will contain the list the
courses and the subjects available. From the list, the
desired subject is selected. Also the different modes
of learning will also be mentioned which makes them
to select their necessary course. The internal tutorial
is the place where the students can find the
availability of notes from various faculty members
and from reputed colleges and also people who have
knowledge about the various courses. The external
links is where there is the availability of various
types of resources namely the PDFs, videos related to Fig. 1. The System Architecture of our
all the courses present both online and offline ontology-based learning support system.

SEMANTIC WEB:-
III. STRUCTURE of ONTOLOGY
Semantic web is the main part of the system where
the search process takes place. It contains parts such
MODEL
as learning XML, learning paths, learning servers and
metadata. Metadata consists of various data about the The hierarchical structure of subject ontology is
data that is present. The search process by the depicted in Fig 3. Subject ontology is constructed
semantic web includes making search with the from one or more of learner-based ontologies,
feature of more machine interpretability. teacher-based ontology, several domain based
ontologies and learning materials. Teacher-based
ontology contains learning concepts and knowledge
REPOSITORY:- structure based on several domain ontologies.
Also, teacher-based ontology is schema ontology to The user interface present has another important
be referred by learner-based ontologies. function to do namely finding the synonyms for the
Learner-based ontology contains concepts and given keyword. The main reason behind is that the
knowledge structure created by students. When a keyword may not exactly match the terms that are
teacher presents learning subjects, students present in ontologies. The synonyms found is given
investigate the subjects and extract meaningful to the User (designer).The class extractor then
concepts and knowledge structure to create a new processes the relevant ontologies that are in the
learner-based ontology or extend existing repository and extracts the class names. With these
learner-based ontology during their learning process. class names returned the process of searching can be
Subject ontology is described as following returned again until the refined required class is
5-tuples, <C, P, I, RH, RC>. The symbol C, P, I, RH obtained. Since the above search leads to many new
and RC represent class, property, instance, hierarchy keywords on refining the User (designer) should
relation between classes and association between maintain a data set that includes the knowledge about
classes individually. We explain the structure of the domain in a wider sense. The User (designer)
teacher-based ontology and learner-based ontology chooses the most relevant concepts that becomes the
based on above 5-tuples in following some classes in t he class diagram.
paragraphs.

Fig. 2. The Structure of Domain Ontology

Development Model

Entities of teacher-based ontology are classified into

following 3 categories:
❖ Learning Concept – Main topics will be
described in a class for a semester. This
category includes fundamental concepts,
advanced concepts, related concepts,
examples and exercises.
❖ Learning Structure – Learning concepts
organized as a semantic network to describe
knowledge structure of topics. In addition,
learning path and schedule represented in
syllabus added to the learning structure.
❖ Learning Material – Teacher collects useful
resources like web pages, images, audios,
and videos and creates lecture notes using
the resources. These lecture notes have
connections to relevant concepts.

IV. DESIGN FLOW DIAGRAM

Fig. 3. Data Flow Diagram
Any user who needs to search for a topic can input
for a keyword for the domain through the web based
interface of the tool. The User Interface uses an V. EXAMPLE
ontology search engine to retrieve ontology files from
the repository which contains the keyword that is Following instance is the Central Conservatory of
mentioned. The ontology search engine performs a Music - Traditional Music - Curriculum - Solo Case
string search on class names, sub-class names and study. We make semantic description for this case. It
property names using the specified keyword in the has the following information:
ontology files containing various data of various Solo musical instrument has three categories: blow
languages. The relevant ontologies are kept in the instrument, string instrument, play instrument.
repository. → Blow instrument contains flute, sheng, xiao.
→ String instrument contains erhu, jinghu, sihu.
→ Play instrument contains guqin, guzheng, lute. <rdfs:subClassOf
Each has the information about history, play method,
rdf:resource="#blowInstrument"/>
and music score.
In OWL, a concept is composed by </owl:Class>
definitions of classes, object properties, data
properties, individuals and axioms. Class is an <owl:Class
important element in OWL. Object properties reflect rdf:about="#fluteHistory">
the relationship between two class instances.
Individual is the member of class. OWL has a very <rdfs:label>fluteHistory</rdfs:label>
powerful reasoning capability, which is based on <owl:equivalentClass>
class. It provides a mechanism to describe the class
that individual belongs to and the properties that <owl:Restriction>
individual inherits by the relationship between class <owl:onProperty
members and then it can be used to reason. The
network structure of this case is shown in Fig 7. rdf:resource="#hasContent"/>
<owl:someValuesFrom
rdf:resource="&rdfs;Literal"/>
</owl:Restriction>
</owl:equivalentClass>
<rdfs:subClassOf
rdf:resource="#flute"/>
</owl:Class>
<owl:Class rdf:about="#fluteMethod">
<rdfs:label>fluteMethod</rdfs:label>
<owl:equivalentClass>
<owl:Restriction>
<owl:onProperty
rdf:resource="#hasContent"/>
Fig 7．Relations of instances <owl:someValuesFrom

All the classes are the subclass of class owl: rdf:resource="&rdfs;Literal"/>

Thing. We define three subclasses of the root class </owl:Restriction>
owl: Thing: score, musical Instrument, music. And
then we define corresponding subclasses of these </owl:equivalentClass>
classes. Take the “flute” class for example; it is the
<rdfs:subClassOf
subclass of the “blow Instrument” class, which is
expressed by the symbol “rdfs: subClassOf” in the rdf:resource="#flute"/>
source code, and it contains two subclass: “flute
History” and “flute Method”. </owl:Class>
The “flute” class and the related class
defined by OWL are shown in TABLE 3. In order to describe the relationship between the
“music” class and the “musical Instrument” class,
TABLE 1. OWL source code - Definition of the that is “music” is played by “musical Instrument”.
“flute” class and its related class The object property is used, which is expressed by
the symbol “owl: Object Property” in the source code
<owl:Class rdf:about="#flute"> as in TABLE 4.
<rdfs:label>flute</rdfs:label>
TABLE 2. OWL source code - Relationship between </owl:Restriction>
the “music” class and the “musical Instrument” class
</owl:equivalentClass>
<owl:ObjectProperty <rdfs:subClassOf
rdf:about="#isPlayed"> rdf:resource="#music"/>
<rdf:type </owl:Class>
rdf:resource="&owl;AsymmetricPropert
y"/> In addition to describing the classes, we also need to
describe their members, which are individuals as we
<rdf:type generally considered. In this case, “short flute” and
rdf:resource="&owl;FunctionalPropert “long flute” are defined as different individuals. In
the source code, rdf: type is a RDF property that links
y"/> individuals (“short flute”, “long flute”) and the class
<rdf:type rdf:resource= that they belong to (the “flute” class). The element
“All Different” declares different individuals.
"&owl;InverseFunctionalProperty"/> After such ontology is established, if we
want to acquire the knowledge about “flute”, the
<rdf:type
ontology inference machine can help return the
rdf:resource="&owl;IrreflexiveProper information of related knowledge including subclass
“flute History”, “flute Method” and the super class
ty"/> “blow Instrument” and so on. Users can choose
<rdfs:label>isPlayed</rdfs:label> accurate learning content from the information
returned or get further information such as the “flute
<rdfs:domain rdf:resource="#music"/> Music” class by retrieval, or continue to make
<rdfs:range in-depth reasoning.

rdf:resource="#musicalInstrument"/> VI. CONCLUSION

</owl:ObjectProperty>
Active learning is one of methods to obtain
higher learning outcomes of students in class. In this
In the source code, the domain of the object property paper, we propose a method to lead active learning
“is played” is defined as “music”, and the range is and self-leading learning of students. Existing models
defined as “musical Instrument”, so it links the for the e-learning process involves having the
“music” class and the “musical Instrument” class. searching process to be done manually by the user.
The link is also reflected through the definition of the This is concerned with the user’s knowledge in the
“flute Music” class, which expresses the relation that subject. In our project we implement ontology using
“flute Music” is played by “flute” as in TABLE 5. the semantic web that ultimately provides results by
searching along with the machine increasing machine
TABLE 5. OWL source code- Definition of the “flute interpretability.
Music” class Learning or gaining knowledge is one of the
key aspects for a student. So the basis for the creation
<owl:Class rdf:about="#fluteMusic"> or any innovation we do comes only when we learn
<rdfs:label>fluteMusic</rdfs:label> the concepts in depth. This site allows students to
learn the concepts easily in which each concept is
<owl:equivalentClass> well organized based on their classifications. It is
very common that people understand a concept or the
<owl:Restriction>
basis from which it has been derived when given in
<owl:onProperty the form of a flow chart. That is what we have
implemented – ONTOLOGY in E-LEARNING.
rdf:resource="#isPlayed"/> This website is implemented for engineering students
<owl:allValuesFrom to learn their subject.
We applied our method to a class,
rdf:resource="#flute"/> Understanding Data Structure, to evaluate the
effectiveness E-Learning support system. We found
that this system enhances the learning outcomes of
students through interviewing from students, defining
specific outcomes, and comparing outcomes of
students before and after applying ontology to the
class.

VII. REFERENCES
1. Hyun-Sook Chung and Jung-Min Kim,
Proceedings of the International
MultiConference of Engineers and Computer
Scientists, Hong Kong
2. Alexander Maedche and Steffen Staab, Learning
Ontologies for the Semantic Web, Semantic Web
Workshop 2001 Hongkong, China
3. Bert Chen, Chen-Yu Lee and I-Chang Tsai,
Digital Education Institute, 2011 International
Conference on Education, Research and
Innovation IPEDR , Singapore
4. Yu D., Zhang W., and Chen X. New Generation
of e-learning Technologies. In Proceedings of
First International Multi-Symposiums on
Computer and Computational Sciences, 2006.
5. Fabrizio Lamberti, Member, IEEE, Andrea
Sanna, and Claudio Demartini, Member, IEEE,
IEEE TRANSACTIONS ON KNOWLEDGE
AND DATA ENGINEERING.
6. Libbrecht, P. Cross curriculum search through
the Geoskills’ Ontology. In Proceedings of
SEAM’08, 2008, pp.38-50.
7. Shackelford, R., McGettrick, A., Sloan, R., Topi,
H., Davies, G., Kamali, R., Cross, J.,
Impagliazzo, J., LeBlanc, R., Lunt, B.
Computing Curricula 2005: The Overview
Report, ACM SIGCSE Bulletin, 38(1), 2006.