'I get a feeling that we are on a...' [The hands make a gesture to show the stern of a sinking ship].
The Monty Phytonesque images on my inner eye from the title of the CETIS 2010 Conference fade and the jolly music of the ship's band starts chiming in my inner ear as I see them move towards the forward half of the boat deck. The CETIS conference is always an upbeat event, even when the prospects for higher education in UK at the moment are not that bright. For an observer from Norway, EduChaos (another sticky concept from #cetis10) is in the range of 0.2% reduction of research funding. However, the numbers mentioned by British colleagues range from 30% to 80% (reduction in research funding vs. funding of teaching). The consequences are hard to imagine for a visitor, even if the smell of burnt tyres are added. One point I do get, though: It's time for Education to innovate, and not only in Britain. This year's CETIS conference was organised to show some possible directions.
For me chaos is a positive term, even if in its original meaning it points to an 'abyss that gapes wide open, is vast and empty' . Yes, it means utter confusion and bewilderment. But it also means the start of something new and better structured. The design of the conference programme gave some ideas on how (a new) order should be achieved. The dynamic force in a conference is the discussion that takes place among colleagues. Most of the time between the opening and closing keynotes was dedicated to discussions in parallel sessions. The keynote speakers made two different bids on framing the discussion and closing the gap:
Before we delve into the CETIS 2010 discussions let us explore how far we get by considering DIY U vs. DSR U.
Anya Kamenetz is a 30-year old journalist who last year wrote a column called "How Web-Savvy Edupunks Are Transforming American Higher Education" . She obviously hit the nail on the head for the publishers, because since then she has written a book on the same theme, done a TeDx show , got herself a manager to keep track of all her tours discussing the demise of old style universities, and even been asked by the customs when entering UK for her opinion on the students protest against higher tuition fees.
Kamenetz has done her research of the University, "the Cathedral of Rationality" (a title from one of her slides), which has gone through constant development for hundreds, if not thousands of years. In the last century, however, the universities had to adjust to the need for mass education. Kamenetz used the California system, established by Clark Kerr  as an example, with its three-tier structure having a handful of University of California research campuses at the top for the elite, and State Universities and Community Colleges catering for the masses. However, Kamenetz concluded:
It was America's best loved welfare program - to share the fruits of knowledge. However, the master plan is dead. I've done a lot of interviews with university leaders all over the US, and the conclusion is unambiguous: The model ran out of capacity.
The current university model does not scale. Anya Kamenetz illustrated the global demand for Higher Education: 53 million in 2000, 120 million in 2010, 250 million in 2025. In India alone, you would need to build a campus for 2,400 people every second week to cover the demand, which is simply physically impossible. The cost of education and the tuition fees are rising, and it is difficult to cut out the cello in the string quartet or ask the band to play faster. Having established that the universities were economically in trouble, Kamenetz went on to show that they are also struggling to deliver. 'Your drop-out rate from university is lower than ours in the US. But is it good enough? What are you getting for your investment?' she asked and answered herself by this quote from a student: 'I Facebook through most of my classes'. With this description of the sad state of affairs Kamenetz had got her equation where she needed to start the reconstruction of the university as a DIY U:
cost + access + quality = the case for radical innovation.
There are many ways to build a case why we need 'revolutionary transformation' of education (which are the even stronger words Obama uses in his National Education Technology Plan ). The challenge is to get the pieces right when we start constructing the new solution. Here I think Kamenetz got it right, as a starting point to moving away from Educhaos. She had identified the three benefits of Higher Education to be Content + Socialisation + Accreditation. And all three have to change.
For Kamenetz, the solution is to put the word 'Open' in front of all three pillars of Higher Education. The two first are well known; and we have made some progress, as we have been discussing Open Educational Resources, and Open Social on the Net for quite a while. However, the Socialisation pillar might be trickier than the Content one, as it is about pedagogy and what should go on inside or outside the classroom.
The essence of learning is found neither inside nor outside the classroom, neither online nor offline. It's in the flow from lived experience and practice, to listening, researching, and sharing the fruits of your work with a community and back out to the world again. Now that so much high-quality information is available for free -- like the 1,900 courses on MIT Open Courseware -- and platforms to allow people to exchange words, images and sound online are exploding in use, many of us are excited about the possibilities of self-organized education that is pared down to this essence, thus affordable, efficient and accessible. 
What is radical in Kamenetz' proposal is the call for Open Accreditation. Institutions may specialise in assessment, helping students to create learning plans, portfolios, etc, and thus breaking the universities' monopoly on accreditation. When HE institutions begin to lose their power to sort and sift people in society, the professional networks could start to play an important role again. And they can bypass the need for diplomas and offer an alternative route to the workplace. Kamenetz scored with the techies in the conference when she referred to Github  as one tool to keep track of your competency revisions.
The bottom line for this American journalist and writer came down to: The world is more complicated now, which means there is more room for people and new ways to do things. After all, the purpose of the Reformation was not have to submit to one cathedral. Cathedral Us will remain, but you could choose the DIY Us instead.
Mike Sharples' answer to the big challenges of climate change, energy shortage, pandemics, terrorism, cultural tension -and education for an inter-connected world, is not do-it-yourself, but a large-scale and multi-disciplinary design science approach. Sharples started his keynote by expressing optimism and focussing on the big opportunities. He pointed to President Obama's plan  as one example of a good approach that tried to address the R&D challenges at the intersection of learning science, technology and education.
Learning is certainly faced with a new complexity, with new interactions (mediated by technology and between learners, institutions and commercial providers), and with new connections (over distance and between formal and informal settings). But this complexity offers new opportunities - for trans-national learning, massively social learning, mobile and contextual learning and life-long and life-wide learning.
Sharples seemed to be of the opinion that we now have consensus on how we learn. As 'evidence' he points to a Science article from 2009 that draws on psychology, neuroscience, machine learning and education as foundations for a New Science of Learning. He quotes Meltzoff et al. that 'insights from many different fields are converging to create a new science of learning that may transform educational practice' . What we need now is to put this new science into action by taking part in design-based research.
Design-based research will improve educational practices through iterative analysis, design, development, and implementation. It is based on collaboration among researchers and practitioners in a real-world setting. And it leads to contextually sensitive design principles and theories . Sharples underlined that this is no longer one community's attempt to design TEL tools. To work, this approach has to be scalable, interdisciplinary and global. So far, so good. However, when Sharples had tightened the canvas and pulled out his crayons it was as if the air had slipped out of the balloon. For all he showed was two examples of tools with a definite pre-social software look & feel, a group scribble and a personal inquiry learning toolkit. These were two examples of scalable design based on research within Secondary Education that were also being extended to Higher Education. For me, however, it was not at all clear how this approach could bridge the Educhaos abyss.
Sharples' conclusion is that this approach to design science has a transformational vision as it is based on open sharing and scaling of best practice and large-scale embedding and evaluation. From the futuristic Learning Sciences Research Institute of University of Nottingham, at the other end of the pond from where the CETIS 2010 Conference took place, more co-ordinated international research work could make a difference. I have to admit I was not at all convinced. And neither was the person who commented from the floor, saying that while the former director of CETIS, Oleg Liber, had predicted that universities would not be around in 2030, everyone had imagined that would be caused by new and disruptive technology, and not that it might be occasioned by a cost-cutting government.
After the concluding keynote I was struck by the sense of uncertainty about our grasp of what is really happening in Higher Education. Even if we know our technology and have a firm foundation in new learning sciences, the political and economical aspects of universities are still a puzzle to the CETIS conference constituency. The response to these kinds of challenges in the CETIS conferences of the last few years has been to look into enterprise architectures and tools for modelling (e.g. Archi ). This year's themes for the parallel sessions showed an opening to unmapped territory by focussing on open innovations; relationship management; and how to create cheaper, more flexible and effective institutions (a session that continued the discussion of the DIY U ), while also looking at more traditional themes like next-generation content and linked enterprise data in F/HE organisations.
The problem with parallel sessions is obvious: You cannot have it all, and you need a little luck to sample the highlights. So it is like learning in general, you get out of it what you are willing to put in. As for myself, I was attracted to the session on Open Innovation as this promised to develop ideas on open innovation strategies for taking advantage of open content, open standards, open data and open source, and for developing more effective partnerships to implement them.
If the maxim for academics is 'publish or perish', the corresponding 'truth' for Higher Education institutions must be something like 'innovate or be eaten'. So how to make innovation happen? The introductory diagram by Scott Wilson (below) seems to suggest that the solution is to put 'open' in front of every word, as in open content, open source, open standards and open data. Then the problem becomes how to get return on investment in innovation. Because, 'we are not hippies! Well, not all of us, anyway…', as it was stated on one of the slides.
Revisiting the Open Innovation diagram, it is clear that the difficult issues are not related to the innovative power of openness, but the inhibiting constraints of drawing too narrow a circle around the business unit within which we work. It is very easy to come up with a number of cases about openness and how the way we share source code, data and content reduces costs. It is harder to describe the generation of income, especially for HEIs.
One exercise was to break into groups and come up with a business model for each corner in the diagram. How to make the four quadrants meet in Open Innovation? Again the new and more competitive climate in the UK was noticed by the oversea observer. 'I've started to hate this - there is no talk about the greater good any more. I used to love the sector I'm working for… this is changed now.' It is much easier to argue that Open Innovation will serve the society as a whole, than the individual institution, which may recently have employed its own IPR lawyer. Even to academics, it is hard to argue that using the Creative Commons NonCommercial is not very useful if you never plan to earn money on your work. If you do use CC-NC, you are also preventing others from innovating and potentially earning money, which arguably is not for the greater good… The session ended up with an action plan, one item being the explicit identification of areas where institutions are competing or are able to collaborate; and another to identify how Open Innovation may fit with the institutional Business and Community Engagement (BCE) strategy .
I have to admit, after the first day I had had enough of acronyms starting with B for business, and decided to go for a more technical track, safe from concerns about sinking ships. I thought Linked Data would offer optimism and soft jokes like 'I never saw a linked data presentation on a Windows computer before… '. I got the jokes ('I will show you how to do triples in Windows'), but the B word was still with us, now packaged as a 'common (business) information layer within the institution' .
Why should we do linked data in the enterprise? asked Wilbert Kraan of CETIS, who already in his session introduction had noticed that increasingly mature linked data tools could enable dynamic and very diverse data sources from across an organisation to be queried and analysed. The promise is to reconcile local concerns with cross-institutional overview, let the systems be what they are, and do the best we can locally, and then, by linking them, release the power of a reconciled picture to management. The data syntax is pretty much solved, according to Kraan. It is merely a matter of transport. The real problem is not technical, but has to do with pragmatics. We need now to focus on the semantics, e.g., what is a student? What counts as a student in different contexts? Introducing linked data to an organisation brings new kinds of discussions and new colleagues to the table.
What are the drivers for this development? Most of the pressures are felt from finance & costing, followed by student data, information management and performance measurement. However, other interests, like staff data, benchmarking, strategic planning, marketing, research data, BCE and estates are all something to gain by looking into linked data and semantic technologies for the enterprise. After Wilbert Kraan had given his hobby horses a canter round the park, I sensed that the participants were at last beginning to feel some optimism. Views were expressed that we should harness existing systems, discuss real issues (at least give them a URI as identifier), find answers both to local needs and the Senate floor. No doubt, the enterprise would be strengthened!
Dave Flanders, programme manager at JISC, brought me back to reality. We are in a recession now and need to make some money. Managing a university is much worse than herding cats; it is like running a company where everyone thinks they are the CEO and are all trying to take the company in their own separate direction. Therefore, Flanders was of the opinion that we need a clear business case for introducing a linked data approach.
Disambiguation is one clear result of the pragmatic activities Kraan advocated. 'How does disambiguation make me money?' was Flanders' question. He answered by giving three cases. First, clearly knowing that the thing you are calling a duck is a duck is invaluable. Second, lists, lists and more lists. Never underestimate the value of structured lists that you can enable other people to reuse. The third case made Flanders explain the subject-predicate-object characteristics of a triple, and we were able to lose ourselves in technicalities for a while. When the RDF roundtrip was done, we were ready for the conclusion about disambiguation. As is so often the case, it is not a technology problem, but a people problem, and the value lies in taking the opportunity to start sorting out issues that need sorting out. It comes to this: It is a question of speed to market. If your institution is the first with information on your courses… Well, then you are first!
Flanders set the tone for the rest of the presentations that showed that semantic technologies are now mature enough to start gathering at the enterprise data waterholes of the institutions. If you also put the data in the open, you even get them cleaned up, which is a business case in its own right. Wilbert Kraan had gathered a good group of experts that were able to explain the subtleties of RDF, RDFa, SPARQL and the rest of it, and to demonstrate how semantic technologies could be used in university enterprise settings.
Damian Steer spoke about the Research Revealed Project  at the University of Bristol. Its purpose is to examine how institutions and researchers can work with the web of research data. Who holds data, e.g., partner institutions, finance, publishers, funding councils, central systems, personal Web sites, etc.? And how could that information be brought together to form a bigger picture?
Paul Miller, from The Cloud of Data , warned against getting over-enthusiastic about specific technologies. Each has its place. We will run into problems if we try to make linked data the answer to every prayer. He pointed to interesting services being built out there without using big words about semantic technologies, but nevertheless using them, like tripit.com, siri.com, and a number of search engines. They all make an effort with the mess people have already created. Linked data may help people clean up their data because they become more exposed to what they have created themselves. But the main thing is to make something useful of what we have got.
Hugh Davis and Yvonne Howard demonstrated how the University of Southampton has used the semantic web to build efficient services to search people, projects, publications, teaching and much more. Going to Hug Davis' official university page  generates on the fly the list of people he works with, Yvonne Howard being one of them. If you want to have that relationship explained, the RKBExplorer will show that they have 20 relations (one might be special, and 14 co-authored papers). The clue is to give everything a unique id: groups, projects, themes, seminars, presentations, people, roles, publications, interest, locations, edges, courts, modules, semesters, sessions, and much more. The rest is simple. So simple, that converting data to RDF could easily be done in the pub, which seemed to be the preferred place to do mashups in Southampton.
Sean O'Riain of the National University of Ireland  gave a tour d'horizon of Enterprise Linked Data, pointing to a number of both useful architectures and tools. (Links are in his presentation you find at the CETIS 2010 wiki which as usual is a good post-conference resource .) O'Riain identified two barriers against the community buying in to linked data: cost and policies. Who is going to publish? And will the management allow you to do it? However, open data are your own data. If you want to publish them, is your choice. Much of the linked data that are out there, exist behind passwords and firewalls. They are still linked data. It might be even more powerful if some of them were exposed to the outside world, but the choice lies with the organisation.
The linked data model makes it easier to let an enterprise eco-system develop over time. Most important is to start, and to start at the right entry point for your organisation. Tim Berners-Lee  has suggested a 5-star deployment scheme for Linked Open Data, and O'Riain gave a link to the Web that offers good examples of how this scheme should be understood . And if senior management needs further convincing, O'Riain threw in a last argument with reference to his fellow countryman's way to run an airline: Linked data make it your own problem getting aboard the plane on time, printing your boarding pass, etc. You could distribute the innovation throughout your organisation. And, it is to be hoped, beyond.
In the current economic climate everything has to be justified, even attending CETIS conferences. Some years ago another overseas participant, Jim Farmer, commented that the CETIS conferences are 'producing benefits that are recognised and will be worth much more than the costs', e.g., facilitating the transformation of Higher Education . This year's conference might be the last one. At least, that was the feeling that spread through the session rooms and was hinted at in opening and closing speeches by the organisers. My hope is that JISC CETIS will continue to bring together the world-leading community in educational technologies in future CETIS conferences. As the Linked Data session clearly showed, it is not about technology any more, but about people and pragmatics. In which case people should meet and extend their links in the open and friendly manner of the CETIS Conference.
For the last decade Tore Hoel has been working on standardisation of learning technologies in Norway, Europe and worldwide. He is now engaged on European TEL projects and is acting as the vice-chair of the CEN Workshop on Learning Technologies. Hoel is co-editor on the new part of the upcoming ISO standard on metadata for learning resources. His field of interest has been standards governance and the role of standards in the development of learning technologies. Hoel has a background in journalism, publishing (founder of a number of professional journals), ICT consultancy, public relations and information management, as well as in ICT and learning.