Ontology-based Photo Annotation Our Approach Developing Ontologies
暂无分享,去创建一个
an image are limited. Typically, the databases have a semistructured indexing scheme that allows a keyword search but not much more to help the user find the desired photograph. Currently, researchers promote the use of explicit background knowledge as a way out of the search problems encountered on the Internet and in multimedia databases. The semantic Web1 and emerging standards (such as the resource description framework (RDF)2) make creating a syntactic format specifying background knowledge for information resources possible. In this article, we explore the use of background knowledge contained in ontologies to index and search collections of photographs. We developed an annotation strategy and tool to help formulate annotations and search for specific images. We also compare our approach’s performance with two existing Web-based search engine options. The article concludes with observations regarding the standards and tools we used in this annotation study.
[1] Corinne Jörgensen,et al. Indexing Images: Testing an Image Description Template. , 1996 .
[2] George A. Miller,et al. WordNet: A Lexical Database for English , 1995, HLT.
[3] Ian Horrocks,et al. OIL in a Nutshell , 2000, EKAW.
[4] Dan Brickley,et al. Resource description framework (RDF) schema specification , 1998 .
[5] Mark A. Musen,et al. The Knowledge Model of Protégé-2000: Combining Interoperability and Flexibility , 2000, EKAW.