Thesauri such as the Art and Architecture Thesaurus (AAT) provide structured vocabularies for describing art objects. However, if we want to create a knowledge-rich description of an (image of an) art object, such as required by the "semantic web", thesauri turn out to provide only part of the knowledge needed. In this paper we look at problems related to capturing background knowledge for art resources. We describe a case study in which we attempt to construct an ontology for a subset of

more »

... ject descriptions, namely antique furniture, using AAT as well as metadata standards as input. We discuss the representation requirements for such an ontology as well as representational problems for our sample ontology with respect to the emerging web standards for knowledge representation (RDF, RDFS, OIL). Keywords Ontology construction, thesaurus, web standards, image indexing There are a number of different forms that text-based image retrieval can take: Keyword search with free vocabularȳ Keyword search with a closed vocabularȳ Thesaurus-based search, where not only the vocabulary is closed but also hierarchical (broader and narrower terms) and other relations can be taken into account in the search process. The general characteristic of this method is that the query is composed of a (possibly Boolean structured) set of terms. The index usually consists of an unordered set of terms. The indexing and retrieval process can both be supported by tools to browse and select terms from the vocabulary. Such browsers are available for large thesauri such as AAT, LCSH and ICONCLASS.