• No results found

2.2 Intelligent Tutoring System Architecture

4.3.9 Author interface

The teacher’s responsibility as an author is to define the domain concepts and the interrelation between them, the example and the exercises, the answers, hints, and far more. This process includes the insertion of available materials into the repository.

Chapter 5

Course Module

Personalized e-learning is a new direction for intelligent tutoring system (ITS) and adaptive hy- permedia systems AHMS. The task of building a course module is a difficult process and requires time and effort. This chapter introduces an ontology to describe the course module of our PAN- DA.TUTOR system. Thus, developing the course ontology is a step toward creating a shared and reusable adaptive educational system. Also, ontologies permits the retrieval of learning material when the pedagogical module has generated the course. The proposed ontology model is a general educational ontology based on various learning strategies and instruction design theories. Fur- thermore, we have realized a separation between the course Content reusability and the learning objects reusability. In addition, this ontology can be used by the author to construct a course with minimum effort. The authored course later is consulted by the pedagogical module to construct a personalized lesson for the student.

5.1

Ontology and Semantic Web

The term ”Semantic Web” embraces efforts to build a new WWW architecture that enhances con- tent with formal semantics. The semantic web is an extension of the current web in which infor-

mation is given a meaning that can be understood by a computer. Likewise, it realizes automatic fetching and processing of web data. The core of the semantic web is ontology, which is used to explicitly represent conceptualizations.

The Semantic Web stack in Figure 5.1 has been proposed and gradually refined by Berners- Lee [120]. It is supposed to guide us through the process of increasing the level of semantics, as well as always to be updated with the new corresponding web technologies. The first three layers contain technologies that are well known from hypertext web and provide a common syn- tax. Middle layers contain technologies standardized by W3C to enable building semantic web applications and add semantics to the Web. The top layers contains technologies that allow new knowledge to be inferred from the explicitly provided information and to check the validity of the statements made in the Semantic Web.

Figure 5.1Semantic web stack

5.1 Ontology and Semantic Web 91

change symbols identified by their unique resource identifier (URI) or internationalized resource identifier (IRI) based on ASCII Unicode. The next semantic layer is XML which represents the structure of data. Extensible XML is general purpose markup language based on IRI, URI and Unicode. XML allows users to add arbitrary structure to their documents but does not encode the meaning of the structure.

On top of XML the resource description framework, RDF is defined. Simply put, it as language to describe whole resources and represent the meaning of data. Thus, the semantic annotation of data is done by means of RDF1. RDF is a data model language developed by the World Wide Web consortium (W3C). It is an infrastructure that enables encoding, exchanging, and the reuse of struc- tured metadata. It is based on the idea that things being described have properties. These properties have values and resources can be described by making statements. Thus, RDF describes how to make statements about resources, in particular, web resources, in the form of subject-predicates- object expressions which are known as triple in RDF terminology, for example ([subject] ’Java’ [predicate] ’has website’ [object] ’www.java.com’). The semantic annotation data gives the pos- sibility to perform some kind of reasoning. RDF Schema2 adds a new layer of functionality by

allowing the representation of ontologies. RDF Schema (RDFS) is a vocabulary description lan- guage for describing properties and classes of RDF resources, with semantics for generalization hierarchies of such properties and classes. RDFS is based on an object-oriented approach and is used by introducing the notion of the class of similar resources, i.e. objects showing a set of the same characteristics. Resources are then viewed as individuals of some class: Every instance of class is also an instance of it’s super class, from which it inherits the properties.

To summarize, resources on the web are considered to be a set of available web resources, each identified by a URI. Such resources are mainly represented by plain XML descriptions. RDF al- lows a web document to be written in a structured way, using a user-defined vocabulary. However,

1(http://www.xml.com/xml/pub/98/06/rdf.html) 2(http://www.w3.org/TP/PR-rdf-schema/)

both RDF and RDFS suffer from the lack of formal semantics for its resource modelling.

The semantic web resolves this problem by adopting URI for the pieces of information it can re- trieve. Besides it supports relations between these pieces. These identifiers are similar to web page URLs but not necessarily point to a web pages. In addition, these identifiers solves the prob- lem of retrieving data. Moreover; the semantic web can link the individual pieces of data to their conceptual categories, in order to support the reusability of data. This can be achieved using on- tologies by considering hierarchal structures of concepts with different relations between them and by binding the individual or instances of data to those concepts. The ontology provides an explicit conceptualization that describes data semantics. The ontology also provides a common vocabulary, interpretable by machines, to researches that need to share information about a domain, including definitions about a basic concept of a domain and its relationships.

This way, the ontology is the backbone of the Semantic Web, a new form of Web content that is meaningful to computers [121]. Ontologies are part of the knowledge representation field of research. They can assist developers to define objects that exist in an application domain [122]. Ontologies define specifications for the conceptualization and the corresponding vocabularies that are used to describe a domain. Besides, ontologies allow the definition of an infrastructure for integrating intelligent systems at the knowledge level, independent of particular implementations, which enables knowledge sharing. Together with various reasoning modules and common knowl- edge representation techniques, ontologies can be used as the basis for the development of libraries of sharable and reusable knowledge modules [123].

Ontologies are appropriate for describing heterogeneous, distributed, and semistructured informa- tion sources that can be found on the web. Typically, ontologies consist of definitions of concepts relevant to the domain, their relations, and axioms about these concepts and relationships. Several representation languages and systems have been defined.

5.1 Ontology and Semantic Web 93

more powerful ontology language. In addition, OWL is the proposed standard for web ontologies representation, and is built upon RDF and RDF Schema. However, OWL provides additional vo- cabulary and allows a representation of more complex relations such as transitivity, symmetry, and cardinality constrains. In other words, OWL supports more powerful expressive capabilities which are required by real applications. In particular, ontologies make knowledge reusable by describing the properties of classes and relations between classes.

OWL provides the three sub-languages owl Lite, owl DL, and owl Full, which offer on the one hand increasing expressiveness, yet on the other hand increasing computational complexity:

• Owl Lite provides the means of defining classification hierarchies, together with simple car- dinality constraints.

• Owl DL offers maximum expressiveness while retaining computational completeness and decidability. The ”DL” illustrates the correspondence to the field of Description Logics.

• Owl Full provides the full, unconstrained expressiveness of the owl vocabulary, yet without any computational guarantees.

Moreover, it is possible to reason about ontologies by means of techniques that are typical for description logics. Reasoning in the semantic web is mostly reasoning about knowledge expressed in some ontology and the ontology layer is the highest layer of the semantic web tower that can be considered as quite well assessed. Top layers contain RIF or SWRL that bring support of rules. With rules, more reasoning can be done over ontologies in knowledge base and they allow verification of knowledge consistency. This is important for example to allow describing relations that cannot be directly described using description logic used in OWL. SWRL has been used to improve the expressivity of OWL language to allow for rules in domain ontologies and to provide a way of automatic mapping between Owl ontologies. Cryptography is important to ensure and verify that semantic web statements are coming from from a trusted source. This can be achieved

by an appropriate digital signature over RDF statements. The user interface is the final layer that will enable humans to use semantic web applications.