Sense and Sensibility

For the closing plenary Brendan Dawes starts by telling us that love and art don't make sense - to a machine. We make decisions that make no sense at all, and enjoy things that are completely pointless. One of the reasons we have such an emotive reaction to the i-phone is that Apple are not afraid to include fun, pointless things in their design process.

App designers are picking up on the need for technology to have a human element. Brendan highlights 'It's A Clock' which tells the time the way humans do - it's just coming up to 12.3o by the way :-) We need to continuely question accepted practices and methods which means not necessarily always designing the most efficient thing. To quote Hendrix - you've got to know what goes between the notes, not just the notes.

Brendan then takes this concept and applies it to data - the transition of the data is very important. DoodleBuzz is a great example of this. It moves away from the 'click' paradigm of using the web and encourages people to doodle across the canvas (webpage) to pull and sort information in different ways. This concept is taking further forward by the Magnetic North website which allows you to draw to pull snapshots of information about the company rather than just presenting pages with 'cv' like descriptions of the company. This is more typical human to human interaction, rather than human to machine interaction. A site I would recommend to all of you to go and look at - now!

Brendan ends by quoting Muriel Rukeyser - the Universe is made of stories not atoms.

How blogs can drive scientific progress

Adam Bly opens by thanking the Tweeters for keeping him in touch with the conference. Adam, you see, is in New York. We're in Edinburgh. Virtual speakers: a first for UKSG.

Distributing, managing, sharing, visualising data
Adam will speak around the issue of how we can ensure that we are constantly updating and refining our first principles and our technologies in a landscape that is constantly evolving. He illustrates changes in science by showing a picture of the large hadron collider. "It's a reminder of how science has become increasingly global, with dozens of countries (even those in current geopolitical conflict) coming together to collaborate." These huge, global, multidisciplinary collaborations require new approaches to distributing and managing data. Scientists must not only produce, but find new approaches to sharing and visualising, data.

"Open science is what will drive the most profound and robust advancements in the future, and will ensure that science has the greatest potential to affect society for the better."

Adam says we have a social responsibility to make research as available as possible to the world - to scientists, and researchers, across borders - and ensure that the technology that does this doesn't impede progress. (It's nice to hear someone make this argument about the core data, not the finished article - I think the former is much more valuable to progressing research, and doesn't devalue what publishers add.) The challenge is not whether open science is good or bad, but how it becomes scaleable, sustainable and simple to adopt. We need purpose-built software and environments for researchers, so that this extraordinarily valuable community is not spending its time on IT but is instead discovering the next great cure.

Starting over on scientific communications
We need to bed four first principles into this reboot of the system:

• Digital core - we've been moving online for years, and will continue to do so. But if the core isn't fundamentally digital, then we just end up hacking solutions around the edge, and can't create the kind of intelligence that a truly digital core for research promises.
• Free flow of information - economic growth is tied to the abundance of scientific information, and we're on track to increasingly free flow.
• Standards and interoperability - to ensure that as projects progress they can be tied together, so individual scientists are not navigating through disconnected and redundant applications, but bringing disparate pieces together.
• Knowledge from information - using tools like data visualisation to see realtime changes and extract knowledge.
  

ResearchBlogging
Adam uses the example of the ResearchBlogging platform, which enables scientists to communicate with each other and the public, using the simple, open blog as a medium. Scientists can easily tag posts so that they can be syndicated to appropriate outlets. It helps ensure that they are contributing to the scholarly record, and that their work and dialogue are being seen more widely than hitherto (journal clubs etc). He shows an example of a ResearchBlogging feed in the PLoS One site - conversations relating to published papers being immediately syndicated back to the source.

Value of blogs in tracking, funding and planning science research
Some researchers in the Netherlands have done the first bibliometric / webometric study of blogging in science, and have used ResearchBlogging as the subject. They found that blogs are more immediate than traditional academic discourse, and are more contextually relevant than academic literature. They focus on the implications of science. This kind of study enables us to understand where discussions around the web are focussed, to better understand the movements of science and direct policy-making and funding.

Solving organisation underload: rethinking scholarly communications to add new conceptual value

"Open Access is going now," says Jan Velterop. "So I feel I can talk about something else - Beyond Open Access." That 'something' is organisation underload.

Too much of our data is too deeply hidden - we struggle to get the most out of it. Jan suggests the problem is not information overload but 'organisation underload'; a lack of organisational conceptual structures to manage all this information. The information overload aspect is going to increase; think expanding communication mechanisms e.g. blogs, peer-reviewed wikis - and why, says Jan, are these not being initiated by publishers?

He uses water as an analogy for information. When there's just a bit, we take it in (we drink it). When there's too much, we have to devise a means to navigate it - a boat. We need to find ways of presenting knowledge that helps us to do something useful and immediate with it. This means not just publishing articles but creating visualisations of conceptually connected data - "this is where the future lies". And as the communications process changes, we will need to think again about what skills and workflows are required to manage it.

As scientists we have traditionally focussed on the detail; now we complement this with a step back to see the bigger picture of how things are connected. This isn't feasible in the traditional manner of ingesting research, but it's this lateral thinking that produces breakthroughs (revealing new conceptual connections).

Jan talks about Knewco's software that mines data for concepts rather than simply words, and (I paraphrase, but I think the essence is that it) breaks the data down into triples that codify the relationships between lots of different pieces of data. Mapping these connections can reveal a powerful picture but once you scale this to millions of triples there will be redundancy that needs to be removed in order to focus on the valuable connections. This kind of semantic data analysis and mapping can make scientific literature even more useful by helping users find new, valuable sources of information - using the connections between literature to support new browse options in library catalogues and publisher websites.

[I have not done Jan's paper justice and would urge you to check out the wealth of additional interesting comments rippling around the Twittersphere (#UKSG09) - and must acknowledge that if I have managed to grasp his thesis at all it is because this is almost exactly what powers Publishing Technology's pub2web platform, which I have spent a good long time getting my head round!]