Archive for July, 2011

Getting to grips with the OpenART ontology

As part of our OpenART project, our project partners at from Acuity Unlimited have been working on an OWL ontology for describing our ‘London Art World’ dataset. This work is nearing completion and will be released in the near future.

For Paul Young (project developer) and I, ontologies are something of a learning curve and we have found that working the ontology  into a diagram has really helped. The following is our current version, still in draft and complete with some rough notes, but in the spirit of sharing I thought it would be useful to make it available using our very own York Digital Library. The diagram contains entities (classes) and object properties for the ‘London Art World’ dataset.

OpenART Ontology Diagram (Draft 2)

OpenART Ontology Diagram (Draft 2)


Leave a Comment

OpenART and Open Licensing

OpenART Licensing

As part of OpenART we need to decide how to license the data that we will expose about the ‘London Art World’ dataset. We said in the OpenART bid that we would make available the data under the terms of the Open Data Commons PDDL but this needed to be further explored with the project stakeholders and the creator and contributors of the dataset.

Within the Linked Data community there is a general desire to be as open as possible. The LOCAH project have used the most open license possible, the Creative Commons public domain CC0 license, and have successfully gained the support of their data owners from the Archives Hub and Copac.

JISC’s recent Open Bibliographic Data Guide encourages the use of a free and open licence: “Universities should proceed on the presumption that their bibliographic data will be made freely available for use and reuse.” (JISC Open Bibliographic Data Guide and “In the vast majority of circumstances, institutions should use a Creative Commons Attribution License (CC-BY) to encourage reuse of copyrightable material. For collections of factual data, the Open Data Commons Public Domain Dedication and License (ODC-PDDL) should be used.” (JISC Open Bibliographic Data Guide Another useful resource for those considering licenses is the DCC How-To Guide, How to License Research Data (
– this identifies some of the easy-to-miss pitfalls of choosing more restrictive licenses.

For OpenART, the Open Data Commons Licenses seem to be the most appropriate license for our dataset. I offered the project team two alternative approaches.

The first was the Open Data Commons Open Database License (ODbL). This is the most restrictive of the Open Data Commons Licenses, but does still allow for wide re-use and sharing. This applies to the ‘dataset’, not its contents. This would allow us to then license the content separately, as may be needed in future for the content contributed by others. For the core data (contributed by Richard Stephens) and the initial release of data, the complementary Database Contents License (DbCL) would be appropriate, which simply places the same condition on the content as on the database as a whole.

What ODbL allows for is summarized here:

In short, this license allows others to re-use freely, so long as they attribute the source, share under the same license and provide unrestricted open access to derived works. It does not explicitly prevent commercial use, but insists on a public and open version always being made available of any derived works.

The alternative is to use PDDL, the Public Domain license, in conjunction with the ‘Community Norms’. This would place the dataset and its contents in the public domain, with users encouraged to abide by the ‘norms’ of sharing the data in the same way (this has no legal basis, it is a statement of good faith).  This approach is asking our contributors to give up their rights in the data.

What PDDL allows for is summarized here:

The ‘Norms’ are summarised here:

The OpenART dataset is the result of several years work and has involved considerable intellectual efforts. Asking it’s contributors to cede copyright and attribution is quite a leap. ODbL, therefore, would seem the best compromise, offering wide-reuse whilst retaining a link back to those who created the data. The decision to use OCbL is not yet final, but remains the strongest contender. I will update the post when a final agreement is reached.

Leave a Comment

JISC YODL-ING Project recent presentations

Our YODL-ING project partners Steve Bayliss and Martin Dow from Acuity Unlimited, presented some of the excellent work they have been doing for us in the project at Open Repositories back in June. The Presentations are available from the conference web site:
  • Stephen Bayliss, Martin Dow, Julie Allinson. Using Semantic Web technologies to integrate thesauri with Fedora to support cataloguing, discovery, re-use and interoperability, Open Repositories 2011, Austin, Texas. PDF
  • Stephen Bayliss, Martin Dow, Julie Allinson. An integrated approach to licensing and access control in Fedora using XACML and the Fedora Content Model Architecture, Open Repositories 2011, Austin, Texas. PDF

Leave a Comment