My Blog has moved!

Thanks for visiting, but my blog has now moved to a new home at www.davidmillard.org, if you have javascript enabled you should automatically be redirected to the right place, if not then please follow the link directly to my new home page.

Friday, October 2, 2009

Killer Semantics: The Challenges of Linked Data in Higher Education

This week I was at ECTEL - The European Conference on Technology Enhanced Learning. The main event for me was actually the workshop for Semantic Technology in Higher Education (SEMHE), where I was on a discussion panel at the end of the day considering potential roadmaps for semantic technology adoption in HE. The atmosphere was unusually pragmatic for this type of event, with real open mindedness mixed with realism - that meant lots of discussion of pedagogy and the potential of linked data.

The question for the panel was what key challenges lay ahead for Semantic Technologies in Education and what might be the Killer Applications. I drew three things from the day that I used as a position statement for the panel:

1) Semantic Technology is often invisible - one of the commonalities of the work presented was that Semantic Technologies like RDF, RDFa, OWL, etc. were often hidden from users deep in the infrastructure. I've argued before that this is a good idea, and that a common mistake with Semantic technologies is to forget that it is for interoperability and inappropriately push it down to the storage layer (e.g. triple stores rather than databases) and up into the interface layer (e.g. graphs rather than forms). However, this does make it invisible to users - like a brilliant new engine oil it makes things run much more smoothly, but users may be unaware of the advantages.


This left me wondering about whether advocacy is the right goal for this community - rather than searching for the Killer Semantic Application perhaps we just let the engineers get on with it, slipping semantics in quietly behind the scenes without a fuss. This would leave us free to look for low-hanging lightweight applications, and to explore the impact on policy and pedagogy. This leads to the next two points...

2) Linked Data is Not Philosophically Neutral - Semantic Technology, and Linked Data in particular, is all about data interoperability and the power of data integration. But this means that, even in a relatively closed system, it involves transparency and openness that can be very threatening to both individuals and organisations. Linked Data will make it possible for students to compare Universities without the usual spin, it will allow Universities to compare students (or potential students) with cold, hard analytics. Linked Data changes the game.

This all means that Linked Data will be a major challenge to policy makers in institutions, who may not even realise what the technology is enabling before it becomes to late. In the past you could disseminate information about your University and market your courses without exposing your data for analysis. This will soon cease to be the case.

3) Technology Changes the Goals of Education - Open Information (i.e. data on the Web) has been very challenging for Education, we have seen a huge rise in problems with Information Literacy (students knowing good sources from bad sources) and in plagiarism (students cutting and pasting from the web). It is easy to lambast students for these academic crimes, but it misses a core truth. That technology has changed what we need to know and the skills that we must possess in order to get by in life. The value of information has changed, and recall is less important when information is available at the click of a mouse or tap of a finger. The problems of plagiarism are a result of educators trying to teach and assess skills that are not seen as valuable by modern students. What's the best way to write a short factual article on any given topic? Actually, a bit of Googling and Wikipedia cut'n'paste is probably the quickest effective solution.

I'm not arguing that this is a good thing (and I've not argued it before), only that it is a change that we must face (for example, by accepting that simple factual articles are no longer a good measure of student's research and rhetorical skills!). Calculators decreased the value of mental arithmetic in schools, allowing teachers to concentrate on higher level skills and methods, but mental arithmetic has side effects (such as improving mental agility) that were thrown out with the bathwater. The solution is not to ban calculators, but to try and understand how this technology changes things, spot the side effects, and develop strategies for managing the change so that we can put the baby back again.

We're still in the process of managing the change to Open Information, but its worth thinking ahead. What might be the impact of Open Knowledge? What happens when the Web can do the things from higher up Bloom's stack - comparison, synthesis, analysis? What happens when you can go to Google and not only ask 'Show me stuff about the English Civil War' but 'How does the English Civil War relate to American Independence?'

So my final suggestion for a challenge is a pedagogical one - what should we be teaching students who will work and study in such a world? What core intellectual skills will society need, and what problems will this technology introduce that will need to be overcome?

Technology has profoundly impacted on education, but the key technologies are not actually those that are specifically about learning. Email, the Web and WiFi have had a huge impact, far above things like QTI or LOM. Linked Data and Semantic Technologies will change society again, and I believe that this contextual change will impact learning far more than specific tools or killer applications.