Object and context profiles

Describing cultural heritage collections is, as most archiving, a job that poses many dilemmas. Traditional collection registration using legacy systems relies on datamodels that are eloborate enough to encompass a number of obvious but different perspectives from which a curator could want to describe the object. The contradiction lies in the notions of "elaborate enough" and "different perspectives".  The famous and notorious Dutch "historical object record" illustrates this very well:

• On the one hand, for practical use the model it is far too big, more then hundred and twenty fields of which in daily practice not more than fifteen, twenty are used.
• On the other hand there is always the complaint that the model doesn't cover enough perspectives. Education for instance is a notorious gap.

A ideal way to describe cultural objects, or any object, would be to be able to dress it up, at random and unlimited, with any possible property  which comes into view. But such a cloud of properties, the complete opposite of the classical fixed record model, poses some severe practical and technical problems.

In the "Groningen casus" of the RNA-project we have implemented a solution which lies somewhere in between those two opposites.

The museums of Groningen, an area in the north of The Netherlands, cover various knowledge domains. Together they manage an impressive amount of objects ranging from farms and old churches to agricultural tools, gravestones, paintings and books. They want to present this varied collection, which digital representations are scattered over a number of database and websites, via the web to a wide audience with search perspectives like cultural history, environmental history, education, tourism, etcetera.

In the Groningen situation we have identified two types of object descriptions, sets of properties that we have called profiles:

• There are profiles which are aimed at describing specific types of objects, like buildings (e.g. farms, old churches), museum objects (e.g. paintings, books, agricultural tools, gravestones ), persons (e.g. persons who lived in the farms, are buried in or near the churches, are mentioned on the gravestones), etcetera. This type of profiles we have called object profiles.
• There are also profiles which are less attached to specific objects, but have more to do with perspectives form which those objects can be observed, like cultural history, environmental history, education, tourism, etcetera. This type of profiles we have called context profiles.

The idea is that any object can be described according to the following rules:

• An object can have unlimited number of object profiles attached to it, but not more than one instance of any kind of these profiles. So a church can be described by only one building profile.
• Object profiles are created and managed by the owner of the object. This includes (1) attaching the profile to an object and (2) filling in the data.
• An object can have more than one context profiles attached to it. So a gravestone can be linked to the educational profiles "primary school", "Warffum primary school, 4th grade", etcetera.
• Context profiles can be created and managed by anybody who is entitled to do so by the owners of the knowledge network. This includes (1) describing the profile and (2) attaching it to objects which are part of the context.. Each context profile has an ID of the body which created the profile.    

The experiment to test this setup in practice has just begun. The museums are linking their objects to object profiles. Next to these they also use context profiles, like "Roman architecture", "agricultural tools after mechanisation",  "art deco", etcetera. At this stage two educational organisations have joined the experiment, they will make their educational context profiles and link them to the museum objects. And the Groningen museums expect that soon other organisations will follow, for instance with tourist context profiles like "museums with restaurant", or "sites on bicycle tour 66", and so on.