Just like other memory institutions, libraries will have to play an important part in the Semantic Web. In that context, ontologies and conceptual models in the field of cultural heritage information are crucial, and the interoperability between these ontologies and models perhaps even more crucial.

CIDOC CRM

CIDOC CRM (Conceptual Reference Model) is an object-oriented model developed under the auspices of ICOM CIDOC (International Committee for Documentation of the International Council of Museums). It was not primarily designed as a data model but, instead, it focuses on the semantics underlying various data structures — it has been rightly labeled as a kind of "semantic glue" for databases, regardless of however different their respective structures can be. It was published in 2006 as ISO 21127:2006.

The CIDOC CRM has its own Web site and several workshops and symposia have been devoted to it and to interoperability among documentation structures for cultural information.

ABC

ABC is a model that has been developed within the Harmony international digital library project to provide a common conceptual model to facilitate interoperability between metadata ontologies from different domains. ABC builds on the core entities from FRBR Group 1: work, expression (though it does not seem that the FRBR distinction between work and expression has been kept in ABC as is), manifestation, and item. In contrast to FRBR, ABC strives to model Temporality through 3 important subclasses of that crucial notion: Event, Situation, and Action. A paper by Carl Lagoze explains how library catalogues could benefit from such an "event-awareness".

There is no formal cooperation between the FRBR community and the ABC community, but there is an effort towards a harmonization of the ABC and CIDOC CRM ontologies. Martin Doerr, Jane Hunter and Carl Lagoze have given a judicious and penetrating criticism of the FRBR manifestation notion in their collective paper Towards a Core Ontology for Information Integration.

EAD

EAD (Encoded Archival Description) is not a reference model such as CRM or ABC, it is rather a data structure standard and encoding scheme for the encoding of electronic archival finding aids. It maintains the multilevel and hierarchical nature of finding aids and facilitates structured presentation and searching of these tools in networked environments such as the World Wide Web. As there are more and more libraries that use the EAD DTD for their manuscript collections, it is extremely important to ensure overall interoperability between FRBR and EAD.

XOBIS

XOBIS (XML Organic Bibliographic Information Schema) may look like a kind of UFO… This project is being developed by the Lane Medical Library at Stanford University. To date, it is the most innovative and revolutionary development in the field of cataloguing theory. It is the most radical effort to get rid of MARC limitations and the "traditional" conception of "bibliographic records", and yet at the same time it also is a supreme achievement in authority control, as even qualifiers are controlled through authorities! This schema takes full advantage of XML linking facilities for library catalogues. Its aim is "to describe the full set of library information, but reorganize this information into a structure that empowers the use of library data as just one more information resource available in the digital domain." Just like in FRBR, the work level is regarded as a core entity in XOBIS. We have much to learn from that visionary project; the FRBR Review Group was very much honoured that Dick Miller, one of the originators of XOBIS and of XMLMARC, accepted an invitation to give a presentation about it on the occasion of the Group's meeting in Berlin on August 4th, 2003. The text of his presentation is available on the Medlane Project Web site.