Philipp Cimiano, Institut für Angewandte Informatik und Formale Beschreibungsverfahren - AIFB,
Universität Karlsruhe (TH)
On the relation between ontology learning, engineering, evolution and
In this talk we discuss the connection between ontology engineering,
learning and evolution of ontologies over time in relation to the
expressivity of the underlying logical formalisms.
While expressive ontologies are needed for a number of applications and
reasonably expressive logics are currently considered within the Semantic
Web, people rarely make use of expressive logical features when modeling
ontologies. In fact, most ontologies are loosely modeled and lack conceptual
preciseness. This is where ontology learning techniques which learn
expressive primitives on the basis of textual data can support a human
As the world and data are constantly changing and evolving, ontology
learning needs to be dynamic in the sense that it is able to detect and
perform necessary changes, thus allowing the ontologies to evolve over time.
However, as we are considering expressive logical formalisms, the process of
ontology evolution can introduce inconsistencies which need to be resolved
in order to allow for meaningful inferences to be derived.
The talk will feature short introductions to the Semantic Web and the field
of ontology learning. Further, we will discuss various approaches for
learning expressive ontological primitives, such as disjointness axioms,
complex class descriptions, rules etc. In order to resolve potential
inconsistencies introduced by such methods, we present an approach based on
finding minimally inconsistent subsets which is guided by axiom metadata.
Finally, we discuss how ontology evolution and other methods for learning
ontologies have been implemented in our own ontology learning framework
Kyo Kageura, Library and Information Science Course Graduate School of Education, University of Tokyo
Terminological Lexicons and Terms in Context: The Translator's Perspective
In this presentation I will try to clarify the relationship between
terminological lexicons and terms in context, or between terminological
lexicon-based research and corpus-based research, from the point of view
of translators' practical need for terminology-related information during
the translation process. This is, in a sense, a variation of the heavily
discussed dichotomy between langue and parole, between language and texts, and
between dictionary entries and language use in actual contexts, etc.
The aim of this presentation is to make the issue clearer and more concrete by
introducing the translator's perspective, which will enable us to set a practical
research agenda as well as to delve into the theoretical issues that reside
in this dichotomy.