Linking research & learning technologies through standards

We have posted about the fluidity of researcher identity, and approaches to identity which acknowledge that fluidity—the NicNames project’s in particular. That post discussed the profusion of identities authors now have online, and presumed that those identities need to be deduplicated, and gathered together so that all the author’s work can be aligned to the one identity—even if we do not presume a notion of primary identity.

But the researcher does not always want their disparate identities tethered together. The pseudonym has long been a literary convention, dissected by literary historians (and authority files). Now it is a mainstay of the blogosphere, where a far amount of scholarly writing takes place; and people are well-attuned to the distinction between pseudonymous and anonymous writing. Internet sleuthing can work out the connections between online identities, just as literary scholars have been doing. That doesn’t mean the authors appreciate if you do. There may be an objective reality about an author’s identity, beyond the fluid consensus of authorities. But fluidity may suit the author just fine, because authors want control over their own identity.

We have mentioned NicNames as an approach to dealing with multiple author identities. The other initiative to mention is an outcome of the UKOLN/DRIVER workshop on international repository infrastructure, held in March. One of the infrastructure tasks the workshop faced was how to establish interoperability between repository identifiers internationally, whether they be identifiers for repository objects, or for authors. At a basic level, repository identifiers from the various available schemes—URL, Handle, PURL, XRI—are already interoperable, since all of them are usable under HTTP. But interoperability is a real problem when it comes to what representations the identifiers resolve to, or how to get a service to operate on identifiers from a huge number of different schemes.

Outside their associated services, though, identifiers are just names associated with things, and the workshop came up with a simple solution to identifier interoperability—which ANDS will take the lead in implementing, as presented at the OAI6 workshop in June. The solution is, have authorities assert that two identifiers are pointing to the same thing. This will allow you to translate queries involving one to queries involving the other, without having to build an extra service layer on top of the existing identifier services.

For author identifiers in particular, the identifiers will be the different tokens associated with researchers by sundry identifier authorities—Elsevier and Thomson, national libraries, grants agencies, institutions. And the authorities asserting equivalence between the identifiers will be national hubs (the UN doesn’t yet have the requisite infrastructure). The assertions themselves can be simple RDF statements of equivalence: katherine.mansfield@hogwarths.edu.au owl:sameAs kbeauchamp@unseen.ac.uk .

So the existing identifiers for authors are left alone, there is no unrealistic proposal to substitute them all with a Single Author Identifier. A layer is imposed over these identifiers, to deduplicate them. And that layer is decentralised, to the national level; because that is rather more feasible than a global solution.

A crucial insight is, these national hubs are still accountable to the researchers, unlike the authority file approach. And they will allow researchers to dissociate online identities, if that’s what they want. So if Kath Mansfield does not want the publications of Kate Jackson associated with her, she can get her national hub to assert instead katherine.mansfield@hogwarths.edu.au owl:differentFrom kbeauchamp@unseen.ac.uk. She can do that if the internet sleuthing associating the two identities is wrong. She can also do it, if it turns out to be right: the researcher is still empowered to control the representation of their own online identity.

To some extent. The national hubs are authorities, in the plural, and there may be another national hub insisting they are the same person after all. And that brings us back to consensus-driven wikiality, as we alluded to in the preceding post. There are authorities to assert two identities are the same, and those authorities are necessary to the scholarly process. But the identities of authors are subject to review and revision—just like the research they publish.

Even if *they* know who they are.

We have already mentioned that the recent quarterly IMS meeting concentrated on developments in Common Cartridge, and how Common Cartridge is being aligned with other initiatives underway in IMS. One of those initiatives is Learning Tools Interoperability (LTI), and was the subject of a developer workshop there.

The other major initiative involving Common Cartridge are Curriculum Standards, which are being added to Common Cartridge as metadata. We have also discussed here the importance of machine-readable curricula, and how they can be exploited as metadata for learning objects—to enable more focused discovery of learning objects, and better alignment of resources to a school’s curriculum. Including Curriculum Standards in Common Cartridge addresses these concerns expressly.
Read the rest of this entry »

Establishing a National Curriculum in Australia has proven to be an elusive goal for the past forty years, for a variety of reasons. The Federal Government has decided to go ahead with establishing a National Curriculum, and has established the National Curriculum Board, to make it a reality [EDIT: now under the Australian Curriculum, Assessment and Reporting Authority (ACARA)]. The Board’s work is well underway: the Shaping Papers and Framing Papers for the curriculum, in the priority areas of English, maths, the sciences and history, have already been subject to stakeholder feedback and revised. The Board is now embarking on the work of scoping, sequencing, and filling in the curricula proper, and aims to publish the curricula after national consultation, in June through September 2010.

Although there has not been a national curriculum in Australia to date, there has already been agreement among the States as to the broad goals of national education, as affirmed by MCEETYA; these include the Melbourne Declaration on Educational Goals for Young Australians in December 2008, and the 1999 Adelaide Declaration on National Goals for Schooling in the Twenty-First Century before it. The National Assessment Program, benchmarking student performance in Literacy and Numeracy, has become nationwide in 2008, but national benchmarking of State tests has already been in place for a decade before that. Independently, MCEETYA has adopted nationwide Statements of Learning for English, Maths, Science, Civics, and ICT since 2003, and these have served to bring the State curricula into at least some alignment. So there are already foundations in place for the National Curriculum Board to build on.

Curriculum objectives and outcomes are stated in prose paragraphs. The entire K-10 or K-12 curriculum can end up being quite a weighty book, given the detail to which a student’s thirteen years of education need to be laid out, and even the broad Foundation Statements . By way of comparison, take the examples of mathematics:

• NSW, K-6
• VIC, P-10
• QLD, 3-9

Even the broad, Foundation Statement summaries of desired outcomes are groups of multiple paragraphs per level. Compare:

• NSW Foundation Statements
• VIC, Standards Table [DOC]
• QLD, Assessable elements and descriptors

But as teaching is transformed in the Information Age, the paragraphs need to be leveraged to maximise their use, beyond the realm of PDFs. In particular, teachers will not want to be guided by the curriculum just to plan what they will teach this semester. They will also want help in working out what resources they can use, to teach those curriculum objectives. With the quantity of resources available to them, and increasing all the time, teachers do not expect to have to sift through the descriptions of the resources one by one, to work out which fits the curriculum objectives best. They quite reasonably expect someone to have done so for them already, wherever the resources were registered and their metadata crafted. Given a curriculum objective, they should be able to search for resources which match it.

The expectation that curriculum objectives should drive resource discovery holds, whether there is a national curriculum, or lots of different state curricula. A single set of curriculum objectives makes things much easier for realising such discovery, because resources need to be searchable against only a single set of objectives, rather than eight. But the most critical requirement for resource discovery is not to have a single set of curriculum objectives per country: there is no national curriculum on the horizon for the States, to take the most obvious example. The critical requirement is to have the curriculum objectives be machine readable. Once the paragraphs of learning objectives can be distilled into assertions to tag content with, content can be discovered through those tags.

If those assertions are machine readable, then the capabilities of the Semantic Web and Natural Language Processing can be brought to bear, to automate the tagging of content as much as possible. Making the tagging efficient is in itself a powerful argument for a consistent national curriculum, since it needs only deal with a single authoritative set of assertions about what students should learn. The efficiencies to be gained are obvious for the publishers and registry providers, who make the content available, as well as the teachers and administrators looking for the content.

Machine readable curriculum objectives have been a long running concern of the e-learning community. The disparity between State curricula, and the need to make content searchable against each state’s objectives, led to the development in 2003 of a Curriculum Organiser component, incorporated in the Basic E-Learning Tool Set (BELTS) tool by The Le@rning Federation. The challenge has been particularly acute in the US, as seen, and has led to initiatives like the Gateway to 21st Century Skills and the Achievement Standards Network. The ASN contains, as a resource, all 51 US curricula and several national curricula in machine-readable form. Based on that resource, members of the Gateway network can build tools to better navigate the assertions in resource discovery and learning pathways, as well as curricula. This leads to the paragraphs becoming active and flexible building blocks: they can be sequenced differently, broken down differently, and matched across streams to the same content, leading to more creative and responsive learning.

As the National Curriculum sets about its work of creating a new set of assertions, there is an opportunity to make these assertions dynamic, integrating them into the ICT-driven workflows that enable 21st century learning. This is especially timely with a national curriculum, which can drive the efficiencies required in those ICT-driven workflows most effectively.