OXLOD/Heritage (Oxford Linked Open Data/Heritage)
Proposal to the University of Oxford's Digital Strategy Group
· Linked Data/Semantic Web is the future.
Using standard web technologies the Linked Data method goes beyond the ‘old’ web, with pages humans read, to the ‘new’ semantic web, with pages other computers can read automatically to enable data in different places and of different types to be linked, enriched and searched in ways previously unimaginable. The potential for academic research, teaching, and public engagement is great.
For example, an object in the Ashmolean Museum is mentioned in a text in the Bodleian Library, restored or analysed for authenticity in the Ashmolean’s Conservation Studio or the Research Laboratory for Archaeology and the History of Art. It may have been discovered and/or acquired by someone listed in the Dictionary of National Biography with rich documentation about the person’s life that stimulates further research. It may have been discovered at a place now precisely located with geo-coordinates and recorded with past/present names in GeoNames, also stimulating new paths of research. It may have been captured in photographs on Flickr, or video on YouTube, with images not available from any other source. And for the wider public automatic links to familiar web resources such as Wikipedia can be useful and stimulate personal research. Enabling these associations transforms research. It makes scholarship accessible and relevant. It also has practical applications in the heritage sector with significant social and economic benefits.
OXLOD/Heritage will make the University of Oxford the first to do that and it will be remarkable for linking data in Academic Divisions to data from Museums and Collections, thus serving scholars, curators, librarians and the public. Scientific data about the analysis and conservation of tangible heritage are also abundant in Oxford (for example, in OUCE/Geography, RLAHA/Archaeology in addition to museums and libraries) and that gives the university additional strategic advantage. Since this aggregated expertise is easily transferable to other disciplines outside Heritage, OXLOD is integral to the University’s Digital Strategy.
· As the pace of technology and globalization accelerates the University must welcome and embed the ‘new’ web. It must be agile and its digital vision must be BIG. It must be global and serve teaching and research. It must also engage with the public and provide leadership in applying digital technologies to scholarship.
· Heritage as defined in this proposal is both tangible (for example, sites, monuments and objects in collections) and intangible (for example, literature and music, ‘traditions’ and ‘performances’, etc.). It embraces all four of the University’s four Academic Divisions (Medical Sciences through History of Medicine) and its Museums and Collections. It is an ideal multi-disciplinary platform for research and teaching. It is also excellent for public engagement because the human past can have interest for everyone and can be understood by everyone. A model developed for heritage is, therefore, directly transferable to other disciplines.
The University’s position
· The University is ideally placed to become a centre for Digital Heritage – broadly defined here as applying digital technologies to the study of the human past, from remotest antiquity to yesterday.
· It is already known as a leader in applying Linked Data in Europe through the Fell-funded CLAROS project (http://clarosnet.org), which is based in the University’s e-Research Centre/MPLS, and in promoting inter-disciplinary and cross-divisional research through the Fell-funded Cultural Heritage Programme (http://culturalheritage.ox.ac.uk), which is also hosted by OeRC.
· That Programme has led Digital Cultural Heritage India and Digital Cultural Heritage China (http://culturalheritage.ox.ac.uk/dch) introducing Asian colleagues to the ISO heritage ontology CIDOC CRM (http://cidoc-crm.org) and W3C standards for Linked Data (http://www.w3.org/standards/semanticweb/data). Its research collaborations have already begun in India with the American Institute of Indian Studies, Archaeological Survey of India and the City Palace Museum of Udaipur, and in China with the National Museum of China, the Dunhuang Academy and the CADAL Digital Library.
· The University can link its academic research, museums and libraries in ways that other institutions cannot because they belong to one legal entity, The Chancellor and Masters. It also has highly experienced units for Research Services and Legal Services to facilitate collaborations.
· It has the technical infrastructure, the cyber security, policies for accessibility and preservation, research expertise in digital technologies, arguably the largest body of digital heritage research of any institution anywhere, and a global brand name that inspires trust. It represents excellence and continuity on the world stage, invaluable in any discipline and critical in heritage.
· Exploit the potential of the Data Cloud (http://www.w3.org/wiki/LinkedData)
· Use heritage as a Linked Data Model for the University’s exemplar digital resources
· Show benefits of opening, linking, and sharing research data across disciplines by ‘doing’ rather than ‘talking’. Assessing the model in open fora will enrich it and promote collaboration. OXLOD is a practical demonstrator of semantic web concepts using an epistemological system for representing knowledge that is independent of a specific technology.
· Show how Linked Data can advance scholarship and promote engagement locally, nationally and internationally across diverse sectors. Data about physical objects can, for example, be linked to scientific/conservation data, to objects discussed in texts, to people and places, etc. This contextualization enriches knowledge. It also makes scholarship relevant and accessible to the public.
· Show how using open source software and semantic web technologies cuts costs by replacing data siloes with data clouds, removing fragmentation, duplication, redundancy, and stimulating global research collaborations.
· Build the model with exemplar datasets across Humanities/SS and Museums and Collections
· Build on existing Linked Data expertise in the University’s e-Research Centre, Computer Science, and elsewhere. Enrich that expertise through existing collaborations within and beyond the University.
· Build a solid digital foundation for the future that can be used across the University for teaching and research, engagement and impact.
OXLOD uses CIDOC CRM (http://cidoc-crm.org), which was created at the behest of UNESCO for ICOM (International Council of Museums) nearly twenty years ago, and has been ISO registered for almost ten. The CRM is well established, broad-based and capable of handling different types of data from different disciplines across the Arts and Sciences. It can also handle both tangible and intangible heritage. Its SIG (http://www.cidoc-crm.org/who_we_are.html) is open, transparent, and internationally highly regarded. The ontology is supported by, for example, the European Commission, Research Councils, and other funding bodies.
To achieve a balance of content, to highlight representative challenges, and to achieve results within one year OXLOD will use sample data. Providers familiar with Linked Data will be encouraged to provide more/all data. Those new to Linked Data will feel more comfortable assessing the potential with sample data. In consultation with colleagues over the past year the proposer (a research academic in Humanities and Social Sciences now based in MPLS) has identified an initial group and secured agreements. Content providers retain their data, images, formats and IPR. OXLOD will not affect their existing databases or websites. OeRC will maintain an indexing cache for searching, not data and images.
Initial sample data will come from:
· Academic research:
o Social Sciences
· Academic coordinator – Professor Donna Kurtz with Professor Dave de Roure/OeRC
· Technical coordinator – Dominic Oldman, Wolfson College, with Professor Dave de Roure
· Programmer (please see profile above)
· Advisory Board (TBC)
· OXLOD evaluates and implements concepts of opening, linking and sharing data
· An estimated 200,000 digital records and images from the University’s academic divisions, museums and libraries will be linked and mapped with an ISO registered ontology to more than 4,000,000 already mapped to that ontology in the British Museum/ResearchSpace, Rijksmuseum and the University’s CLAROS, the latter with data from more than ten content providers in six European countries. This exemplar will be without parallel.
· Content providers from the University’s Academic Divisions, Museums and Libraries will meet regularly to discuss common challenges while gaining knowledge of Linked Data
· Technology experts from OeRC, OII, Computer Science, Engineering Science, SBS and others will be encouraged to work together on Linked Data research. Bringing this expertise together and applying it to digital resources will provide another exemplar without parallel.
· Different technical approaches to Linked Data will be assessed to provide a firm foundation for future development
· Bringing Linked Data to academic researchers, curators and librarians will strongly enhance their ability to leverage funds from internal and external sources
· Embedding Linked Data expertise in the University’s e-Research Centre will build a foundation for future digital research and maximize the potential of cross-disciplinary collaboration