Building And Evaluating Domain Ontologies PDF Download

Ontology has been a subject of many studies carried out in artificial intelligence (AI) and information system communities. Ontology has become an important component of the semantic web, covering a variety of knowledge domains. Although building domain ontologies still remains a big challenge with regard to its designing and implementation, there are still many areas that need to create ontologies. Information Science (IS) is one of these areas that need a unified ontology model to facilitate information access among the heterogeneous data resources and share a common understanding of the domain knowledge. The objective of this study is to develop a generic model of ontology that serves as a foundation of knowledge modelling for applications and aggregation with other ontologies to facilitate information exchanging between different systems. This model will be a metadata for a knowledge base system to be used in different purposes of interest, such as education applications to support educational needs for teachers and students and information system developers, and enhancing the index tool in libraries to facilitate access to information collections. This thesis describes the process of modelling the domain knowledge of Information Science IS. The building process of the ontology of Information Science (OIS) is preceded by developing taxonomies and thesauruses of IS. This research adopts the Methontology to develop ontology of Information Science OIS. This choice of method relies on the research motivations and aims, with analysis of some development ontology methodologies and IEEE 1074-2006 standards for developing software project life cycle processes as criteria. The methodology mainly consisted of; specification, conceptualization, formalization, implementation, maintenance and evaluation. The knowledge model was formalized using Protégé to generate the ontology code. During the development process the model has been designed and evaluated. This research presents the following contributions to the present state of the art on ontology construction; - The main achievement of the study is in constructing a new model of Information Science ontology OIS. The OIS ontology is a generic model that contains only the key objects and associated attributes with relationships. The model has defined 706 concepts which will be widely used in Information Science applications. It provides the standard definitions for domain terms used in annotation databases for the domain terms, and avoids the consistency problems caused by various ontologies which will have the potential of development by different groups and institutions in the IS domain area. - It provides a framework for analyzing the IS knowledge to obtain a classification based on facet classification. The ontology modelling approach is based on topdown and bottom-up. The top-down begins with an abstract of the domain view. While the bottom-up method starts with description of the domain to gain a hierarchal taxonomy. - Designing Ontocop system a novel method presented to support the developing process as specific virtual community of IS. The Ontocop consists of a number of experts in the subject area around the world. Their feedback and assessment improve the ontology development during the creating process. The findings of the research revealed that overall feedback from the IS community has been positive and that the model met the ontology quality criteria. It was appropriate to provide consistency and clear understanding of the subject area. OIS ontology unifies information science, which is composed of library science, computer science and archival science, by creating the theoretical base useful for further practical systems. Developing ontology of information science (OIS) is not an easy task, due to the complex nature of the field. It needs to be integrated with other ontologies such as social science, cognitive science, philosophy, law management and mathematics, to provide a basic knowledge for the semantic web and also to leverage information retrieval.