Semantic Web-Based Information Systems: State-of-the-Art ...

Semant c Web of Language Eng neer ng
for annotation; (3) the management of ontological tiers, which can be annotated
with language profile terms and, therefore, corresponding ontological terms; and
(4) storing OntoELAN annotation documents in XML format based on multimedia
and domain ontologies. To our best knowledge, OntoELAN is the first audio/video
annotation tool in the linguistic domain that provides support for ontology-based
annotation. It is expected that the availability of such a tool will greatly facilitate
the creation of linguistic multimedia repositories as islands of the Semantic Web
of language engineering.
Introduction
The Semantic Web (Lu, Dong, & Fotouhi, 2002; Berners-Lee, Hendler, & Lassila,
2001) is the next-generation Web, in which information is structured with well-defined
semantics, enabling better cooperation of machine and human effort. The Semantic
Web is not a replacement, but an extension of the current Web, and its development
greatly relies on the availability of ontologies and powerful annotation tools.
Ontology development and annotation management are two challenges of the
development of the Semantic Web, as we discussed in Chebotko, Lu, and Fotouhi
(2004). In this chapter, although we use our developed general multimedia ontology
as the framework and the GOLD ontology developed at the University of Arizona
as an ontology example for ontology-based annotation of linguistic multimedia
data, our focus will be on addressing the second challenge — the development
of an ontology-based multimedia annotator OntoELAN for the Semantic Web of
language engineering.
Recently, there is an increasing interest and effort for preserving and documenting
endangered languages (Lu et al., 2004; The National Science Foundation, 2004).
Many languages are in serious danger of being lost, and if nothing is done to prevent
it, half of the world’s approximately 6,500 languages will disappear in the next 100
years. The death of a language entails the loss of a community’s traditional culture,
for the language is a unique vehicle for its traditions and culture.
In the linguistic domain, many language data are collected as audio and video recordings,
which impose a challenge to document indexing and retrieval. Annotation
of multimedia data provides an opportunity for making the semantics explicit and
facilitates the searching of multimedia documents. However, different annotators
might use different vocabulary to annotate multimedia, which causes low recall
and precision in search and retrieval. In this article, we propose an ontology-based
annotation approach, in which a linguistic ontology is used so that the terms and
their relationships are formally defined. In this way, annotators will use the same
vocabulary to annotate multimedia, so that ontology-driven search engines will
retrieve multimedia data with greater recall and precision. We believe that even
Copyright © 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission
of Idea Group Inc. is prohibited.