Recent Advances in Electrical Engineering and Educational Technologies A Study on Ontologies and their Classification Thabet Slimani ontology has a distinguished naming which depends on the Abstract— During the last decade, the use of ontology in distributed domain: environments (Web) is useful to enable semantic management of • in the field information retrieval is named thesaurus information and knowledge sharing. Given that the objective of an • in the field of linked data is a model represented by ontology is to capture generic and formal knowledge, used and OWL format shared across applications and by groups of people, it is important to know their classes or families which help their content understanding. • in the field of databases is an XML schema In this paper, after a general introduction about the basis of • etc. ontologies and its modularization, we present ontology classes which However, to perform a study clarifying the different forms represent an important issue in semantic web. of ontologies to help their content understanding is an important issue. Additionally, it is important to define precisely the vocabulary derived from the word ontology. For Keywords—Ontologies, ontology classes, ontology families, example, what is the difference between a domain ontology ontology languages, ontology modularization, semantic web. and a task ontology? I. INTRODUCTION First, we introduce the modularization of ontologies. ECENTLY, the word ‘‘ontology’’ has become a stylish Second, we present the different types or families of R word within the Knowledge Engineering discipline. ontologies. Several definitions about ontology have been presented in many papers. Also, we have seen how such definitions have modified and evolved over the time. At least, an ontology II. ONTOLOGIES: DEFINITIONS AND MODULARIZATION allows a machine-understandable semantics of data, and make easy the search, exchange, and integration of knowledge. Which facilitate the use of e-technology. The word ontology is employed in the field of AI research, The following section presents two different definitions of as it is useful to make the conceptualizations of a domain Ontologies: explicit which enables their comparison and analyzes. Gruber • In computer science, an ontology is an attempt to make a [2] defines an ontology as an explicit specification of a complete and rigorous conceptual schema within a conceptualization. An ontology is a description of knowledge- specified domain. Typically, an ontology is a hierarchical level [4] where the representational formalism is independent data structure including all the significant elements and [5]. Additionally, ontology is a representation of the entities their relationships and regulations within the domain [1]. type, their relations, and their constraints [4]. It consists of a • In AI field, an ontology is an explicit specification of a set of classes, relations, instances, functions and axioms conceptualization [2, 3]. In such an ontology, the ordered hierarchically. Formally, ontology is a description of universe of discourse is concept names (e.g. classes, data that remains constant over various data/knowledge bases relations, axioms) accompanied with a description of in a certain domain [6]. what the concepts mean, and their formal axioms. The differentiation between Ontologies based on their generality level can be presented in the classification Although the use of ontologies suggests a concrete approach presented by Guarino in [7] and showed in Figure 3. to build sharable knowledge bases, it also raises questions that Ontologies can be classified according to the need answers, such as: conceptualization subject (content). Very general things such • Which are the different ontology families? as time, space, insubstantial or concrete objects, and so on can • What is the objective for using an ontology in an be covered by the top-level Ontologies, independent to the application? domain usage. The construction of either domain or task • Which approach can be used to build an ontology? Ontologies can be done based on these top-level Ontologies. Several types of ontologies have been enumerated in The first category includes Ontologies dedicated to cover a academia. Depending on context, the word “ontology” can given domain (medical or university, for example) designate different computer science objects. For example, an independently to the task that uses the ontology. The second category includes Ontologies specified for a generic mission (content annotation or situation recognition, for example) T.Slimani is with Computer Science Department, Taif University, Taif, Saudia arabia, on leave from the High School of Technological Study, Tunis, irrelevant of usage domain. In conclusion, the development of Tunisia (e-mail: [email protected], [email protected]). ISBN: 978-1-61804-254-5 86 Recent Advances in Electrical Engineering and Educational Technologies application Ontologies helps particular tasks to be solved - Instance Set: Constitutes a set of instance that within particular domains, and consequently often requires belongs to a concept. For example, “firdous paradise” both domain and task Ontologies for reusability. is an instance of “afterlife-location” concept of Quran ontology. Figure 1. Guarino’s ontology classification [7]. Ontology may be classified as follows, based on the scope of the ontology (see also Figure 1): • Upper/top-level ontology: it describes general knowledge (i.e. what time is and what space is). • Domain ontology: it describes the domain (medical Figure 2. Ontology Depicting Ontology Components with their domain, personal computer domain or electrical relationships. engineering domain). • Task ontology: it is suitable for a specific task Ontology components such as concepts (things, events…), (assembling parts together). instances and properties are represented by one or more • Application ontology: it is developed for a specific symbols denoting terms rapidly understood and readable by application (assembling personal computers). humans. The connection between all these ontology components is realized through relations. Semantic relation is Modularization can be used at each level. For instance, upper a connection that link only concepts together: for instance the ontology could includes modules for real numbers, time, and location relationship indicates that student concept is enrolled space (parts of upper ontology, generally are called generic in a university concept. The link that connects only instances Ontologies). Upper levels Ontologies could be imported by is denoted by instance relations which are in turn instances of Ontologies at lower levels and adding them specific semantic relations. It is difficult to generalize the relation knowledge. Domain and task Ontologies may be independent between all instances of their concept, because some relations and are combined to be used for application ontology. between instances are contextual. As example of instance relation the student instance named “john” is enrolled in the university instance named “Stanford”. An example of III. ONTOLOGY FAMILIES contextual instance relation can be that the professor instance named “Jeffry” is localized in the university instance named “Stanford”. The relationship that terms can have is expressed The classification of Ontologies depends on several criteria. by the terminological relations. For example the term It is possible to make classification based on the ontology “university” is synonym to the term “education”. expressivity (language of representation) or the ontology According the type of structure and the amount of their use, components (concepts, instances, properties, axioms, etc. See Sowa [8] distinguishes two main categories: terminological Figure 2). Ontology components can be represented by (lexical) and axiomatized (formal). specific Ontologies. For example, if we focus on concepts, as The following sections present four categories of Ontologies a key component of Ontologies, they can be represented by an (including the two categories presented in Sowa) ontology based on different behaviors: corresponding to the usage of the previously described - Textual Definitions: As instance it defines concept components. Each section describes the type of language used by a sentence (“University”, “Person”….) to define the ontology. - Properties set: As instance the concept “University” has the property “name”, “address” and “creation A. Informational Ontology date”. - Logical definition: Constituted by several formula. For example, in Quran ontology, the Earth , Sun, and Informational Ontologies is a composition of diagrams and Moon are classified under "Astronomical Body": sketches utilized for the clarification and organization of collaborators project development ideas. These Ontologies can ISBN: 978-1-61804-254-5 87 Recent Advances in Electrical Engineering and Educational Technologies be used only by humans can use these Ontologies. Linguistic Ontologies can be glossaries, taxonomies, Informational Ontologies are characterized by the following dictionaries, folksonomies, controlled vocabularies, thesaurus, behavior: or lexical databases. A linguistic ontology concept can be • Simply adjustable and scalable referenced by different terms (computer, IT and computing are • Artificial and schematic synonyms) determining the relative positions of the concepts • In general, they can be used at some point in the with respect to one another and partially specified by relations