Held in Conjunction with the 19th Australian Joint Conference on Artificial Intelligence (AI'06)
December 5, 2006, Hobart, Tasmania, Australia
Abstract: Mephisto is a framework that will enable ontology-based high-level information fusion. Within the framework, the name Mephisto can be used to refer to a conceptualisation, an ontology, an agent, a society, a formal theory, and an implementation. A Mephisto conceptualisation assumes that the world can be understood in terms of processes. A Mephisto ontology is a specification of the conceptualisation. A Mephisto agent employs the ontology. A Mephisto theory is a formal theory of processes. A Mephisto implementation implements the theory. A Mephisto theory plays a crucial role within the framework.
Abstract: This paper presents an empirical evaluation of description logic reasoners to support the selection of scalable ontology engineering patterns for TBox reasoning. Our main objective is to define the rationale behind the design decisions required for the generation of large ontologies with XSLT-based tools. We discuss here the outcomes of an experiment focusing on aircraft components and parts for which we have implemented the ontology design guidelines for part-whole relationships published by W3C's Semantic Web best practices working group. We have worked with the following reasoners, being the best state-of-the-art currently available: FaCT++, RACER, Pellet and CEL. We found considerable variation in reasoner performance and have attempted to characterise the factors that distinguish the reasoners to enable a bestpractice design style to be successfully applied for the generation of very large ontologies..
Abstract: As ontologies become more prevalent for information management the need to manage the ontologies increases. In the community services sector multiple organisations often combine to tender for funding. When separate organisations come together to generate reports for funding bodies an alignment of terminology and semantics is required. Ontology creation is privatised for these individual organisation to represent their view of the domain. This creates problems with alignment and integration, making it necessary to consider how much each ontology should influence the current decision to be made. To assist with determining influence a trust based approach on author and the ontologies provides a mechanism for ranking reasoning results. A representation of authors and the individual resources they provide for the merged ontology becomes necessary. The authors are then weighted by trust and trust for the resources the author provides to the ontology is calculated. This is then used to assist the integration process allowing for an evolutionary trust model to calculate the level of belief in the resources. Once the integration is complete the semantic agreement between the ontologies allows for the recalculation of the authors trust.
Abstract: One of the most successful Healthcare Information Models is version 2 of the Health Level 7 (HL7) standard. However, this standard has various problems, mainly its lack of semantic interoperability. This shortfall was addressed in HL7 Version 3, a new standard which has been designed to solve this problem. Total semantic interoperability cannot be achieved without defined terminology, and to this end the use of The Systemised Nomenclature of Medicine - Clinical Terms (SNOMED-CT) is proposed. The difficulty arrives when deciding how to integrate the information model and the terminology. The line between where one ends and the other begins is often indistinct. This paper describes a proposal for normalising the two using ontology mapping and basing HL7 message models on SNOMED-CT concepts and their relationships, in an effort to further total semantic interoperability and seamless communication between healthcare entities.
Abstract: Ontologies are widely used in text technology and artificial intelligence. The need to develop large ontologies for real-life applications provokes researchers to automatize ontology extension procedures. Automatic updates without the control of a human expert can generate potential conflicts between original and new knowledge. As a consequence the resulting ontology can yield inconsistencies. On the other hand, even if the information extracted from the external sources automatically is consistent with the original ontology it can be generalized unsystematically and conceptually wrong what will lead to mistakes by the application of the extended ontology. We propose an algorithm that models the process of the adaptation of an ontology to new information and regeneralizes the resulting ontology in a more intuitive way inserting additional knowledge where it is possible.
Abstract: A number of ontology-based approaches have been suggested for the description of service behaviors to be used in service composition and matching in service oriented architectures. We examine an approach based on classical software engineering notation and compare it to other approaches.
Abstract: Ontology matching (or mapping)---finding correspondences between semantically related entities of heterogeneous ontologies---becomes crucial for interoperability in distributed and intelligent environments. Although some efforts in ontology mapping have already been carried out, the issues of semantic heterogeneity and expert-interaction in a mapping process still need to be considered. Our intuition on these issues is inadequate semantics and unstructured taxonomies in ontologies. In order to overcome these obstacles, we propose a semantically enriched model of ontologies (called MetaOntoModel) where the semantics of concepts are enriched by adding domain independent knowledge (called meta-knowledge) based on three philosophical notions: {\it identity}, {\it rigidity}, and {\it dependency}. Our novel idea is if two concepts possess different kinds of meta-knowledge, then they are not possible to be matched. Thus, a direct concept matching is driven between the same meta-knowledge groups of two heterogeneous ontologies and it would reduce the complexity of matching process.
Abstract: During the last years, there has been an increasing interest on ontology due to its ability to explicitly describe data semantics in a common way, independently of data source characteristics, providing with a schema that allows data interchanging among heterogeneous information systems and users. A lot of works have been done to improve ontology technological aspects, like representation languages and inference mechanisms, and less attention have been paid on practical results of development method application. This paper discusses the process and product of an experience in developing ontology for Public Sector which organization requires a strong knowledge managment. Particularly, this process was applied to develop ontology for Budget Domain.
Abstract: This paper discusses the manner in which SNOMED CT (SCT) has confused the metonymic role of some class labels as holonyms and has inappropriately assigned property inheritance down a holonymic chain due to its transitiveness. The notion of emergent properties is introduced as the only form of property that can exist on a holonym and its use in a hypernymic inheritance hierarchy is discussed. The consequences of this modelling approach for SCT are discussed and the use of metonymic substitution for holonyms at the point of care is presented as a source of confusion for causing the modelling of inheritance in holonymic hierarchies for clinical care.
Abstract: In this paper we present a user friendly approach to annotate websites with machine-processable information in controlled natural language. The controlled natural language serves as a high-level specification and knowledge representation language which allows human annotators to summarise individual web pages of a website and to express domain-specific ontological knowledge about that website in an unambiguous subset of English. The annotation process is backed up by an intelligent text editor which supports the writing process of the controlled natural language with the help of text- and menu-based predictive interface techniques. The text editor runs as a Java applet and is connected over the Internet to a controlled natural language processor and to a reasoning engine (consisting of a theorem prover and a model builder). The controlled language processor translates the summaries of web pages and the ontological knowledge about a website into first-order predicate logic and the reasoning engine combines this information into a set of micro theories for consistency and redundancy checking as well as for question answering. Specification texts written in controlled natural language are both human-readable and machine-processable, and can be easily exported and distributed as web feeds.
Abstract: Domain ontologies and knowledge-based systems have become very important in the agent and semantic web communities. As their use has increased, providing means of resolving semantic differences has also become very important. In this paper we survey the approaches that have been proposed for providing interoperability among domain ontologies. We also discuss some key issues that still need to be addressed if we were to move from semi to fully automated approaches to provide consensus among heterogeneous ontologies..