Technical Information (2011)
Since the first generation of major academic resources on the web 10 years ago, there have been many changes in both tools and in user expectation. We have excellent mapping and satellite services from Google and others; the semantic web technologies which let us navigate the web using something other than simple textual searches are starting to deliver on their promises; the approach to writing exciting interfaces known as Web 2.0 has matured vastly; our readers have an an increasing expectation of free access to almost unlimited amounts of data which is immediately useable and linked to other data; and our systems expect to access material with a computer-computer interface, not just humans looking at web pages.
CLAROS, then, set out with five guiding principles in its aim to be a leading resource for the next decade:
- Multi-domain and worldwide: we start with archaeology and art but we want to encompass any aspects of material culture, and want to provide ways to join up material from all over the world from any period
- Cross-disciplinary: we learn from, and contribute to, the concerns of colleagues in e-Research, image processing, database design, and user interfaces
- Semantic: CLAROS data is to be managed according to established ontologies
- Multi-level access: we want to offer as many different interfaces as situations or people may demand, and we will take full advantage of the new possibilities offered by mapping, digital cameras, mobile devices, and modern interface design
- In summary, CLAROS is open: the material is to be useable by anyone for anything
How does CLAROS arrive at unified access to a web of data integrating access to scholarly information on archaeology and art? At the time of writing (2011) we are providing over a quarter of a million records about classical Greek pottery and gems, thousands of antiquarian photographs, details of 400,000 persons with Greek names, several thousand records of eastern art, another quarter of a million records about classical statues, and 5,000 records of mythological and religious iconography; and relating the majority of these records to a set of 8000 geolocated sites around the world.
The CLAROS infrastructure relies upon four assumptions:
- Proportionality and subsidiarity: CLAROS does not attempt to model all of the information about artefacts and places, just that subset which is useful for searching, which the partners are comfortable releasing with an open licence, and which we are confident can be described in the same semantic framework.
- The working practice is one of federation. CLAROS ingests a catalogue of records from each data partner and amalgamates it into a single entity, but for more detailed information about a hit we also return to the original web site of the partner. CLAROS is a resource discovery service, and its job is to provide cacheing, indexing, querying and visualization services.
- We all use a common ontology, a vocabulary for classifiying the data; this is the conceptual reference model developed by CIDOC for describing museum and cultural heritage information. It is widely known, and beginning to be adopted in practical projects by archaeologists and museums. CLAROS is one of the largest collections of CIDOC CRM material available at this time.
- We manage the data using the semantic web standard Resource Description Framework (RDF), managed using a special form of database called a triple store, and queried using the SPARQL language.