Cidoc CRM is a relatively well-known acronym in these circles. In 2003 a joint group was formed to examine how the two (then) emerging information standards, FRBR and Cidoc CRM, could communicate.
It was a joint libraries and museums collaboration, explained Patrick. Of course, museums and their collections are quite different from libraries. Libraries, unlike museums, hold non-unique copies of books. The fundamental cultural content, or originality, or value, or uniqueness of the book is not deemed particularly relevant.
Within the library sector, the key concepts for classifying material are things like authors, titles, topics. In museums, however, it’s all about unique items, about the cultural circumstances under which they were produced and how they might be related. The culture of the individual object is extremely relevant, unlike the mass-produced book.
The project began with Cidoc CRM as a reference point, because at that time, FRBR was not finished. Since then, Cidoc has undergone ‘extensions’ to the initial schema.
Patrick outlined ways forward that came out of the project; FRBR attributes were extracted and semantic opportunities explored as a result. FRBR items/terms as a result may now be able to express their semantic meaning using existing Cidoc constructs. Both communities will benefit from a unified object-orientated model for bibliography and museum info, but the combined model will only show its benefits when the semantic web develops technology more. And of course, such linked networks need good quality data.
Afternoon session, day one
Professor Eero Hyvonen head of the Semantic Computing Research Group (SeCo) at Helsinki University of Technology (TKK) Finland, woke us all from our lunch with the phrase – ‘if you look under the hood of the semantic web there are no certainties.’
He describes this as a national problem. There are often no ontologies at a national level, but the semantic web needs a content infrastructure, like traffic needs roads.
Business applications can be built on solid roads. URI are the point of ontologies – better than keywords. URI’s are universally shared with all users. He sees a future with nested sites, working together semantically. Using intelligent semantic web structures relies upon the deployment of ontologies.
Researchers from ITT at Helsinki then explained some of the cutting edge research happening at the University, a session begun by Kim Viljanen, who outlined work carried out developing ontology servers. Some key questions already arising concern old content or collections. How should ontologies be used in relation to legacy systems?
We heard about the excellent effort in Finland to build agreements to use recommended ontologies across the various digital domains. There’s a media art ontology, a photography ontology, Iconclass – an art and culture ontology, a time ontology – aika, Actor ontology – persons, groups, organisations
ULAN from Getty.
As a result, in Finland there is now a really complete set of relationships with very interesting possibilities. Right now there is a national ontology service- perhaps one day a global ontology service?