#Solo11: Best of the rest
Last weekend I attended this year’s Science Online London conference (which the Wellcome Trust sponsored). This week I’ve been reporting on a few of the key things I took from the conference. This final post rounds up a few of the parallel sessions and workshops at the conference.
- How are wikis being used to carry out and communicate science?
- Can we develop something like Schema.org to encourage data sharing and reuse?
- Bridging the divide: building around the PDF
- So many ways to tell a story
- A National Undergraduate Bioscience Research Journal
- Workshop: Online Communications Tools 2
How are wikis being used to carry out and communicate science?
Many different ways researchers and research organisations are engaging with Wikipedia and Wikis in general to communicate science. But how?
Wikipedian Michael Peel gave a quick overview. There are many ways that the Wikimedia Foundation (parent of Wikipedia) are reaching out: WikiNews, Wikiversity, Wikipedia selection for schools CD (an offline snapshot of articles as a CD-ROM encyclopedia). Elsewhere, BBC Wildlife uses Wikipedia to pull in descriptions of animals. And the Wellcome Trust Sanger Institute Rfam project puts in (and pulls out) descriptions of RNAs using Wikipedia. Meanwhile, museums are using QR codes next to exhibits to allow visitors to find out more – the code can even detect the viewer’s settings and send them to the relevant language Wikipedia page when scanned.
Perhaps one of the most striking examples of Wikipedia use, however, is Cancer Research UK’s recent exploits, which gained unexpected media splash as a result). Henry Scowcroft, News and Information Manager at CRUK, said the big question is not whether we should engage, but how to fit it into our day jobs?
There’s also the problem of politics – how do we deal with reversions of our edits in the constantly-changing Wikipedia? This is a particular issue from an organisation’s point of view — Wikipedia community’s standpoint is that anyone associated with an organisation shouldn’t be pusshing their agenda, but what constitutes ‘bias’ and ‘marketing’, and what ‘expertise’? Several of us have been wondering whether an organised approach learning from the GLAM Wiki initiative could help. In that example, museums, libraries and galleries have gained trust for experts and institutions by working with the Wikipedia community, gaining an ‘authorised pass’ if you will. Perhaps a project based approach is the way forward.
CRUK’s Wikipedia Club meet once a week to edit articles related to cancer. This can be as simple as adding news (including non-CRUK but cancer-related) info to update Wikipedia articles.
Guardian journalist Alok Jha then gave a perspective of Wikipedia as a source. He said it’s not the text that’s useful, but the references. You wouldn’t cite it as a source, but it provides a useful starting point. It’s good for discussion around a topic, books, reports, papers, primary sources (long before we had things like Twitter to provide sources and leads).
He also made this important point: we respond to stories not data - Wikipedia gives us that, which is maybe one reason why it’s a popular way for us to get information.
Jha said it is a great way of filling in gaps — particularly on things like movie or TV plots. It’s an interesting point that Wikipedia is used widely by most people in their personal lives, but often not in a professional context. He pointed out that there are lots of science-related Wikipedia articles but they are very hard to understand (“garbage to me”). They assume too much knoeledge and don’t ‘teach’ science very well – but is that what it’s trying to do?
Can we develop something like Schema.org to encourage data sharing and reuse?
This session, aiming to answer the above question, was led by Dr Tomi Kauppinen from the Institute for Geoinformatics at the University of Münster. In case you are unfamiliar with Schema.org, it provides a collection of schemas, i.e. html tags, which webmasters can use to mark up their pages in ways recognised by major search engines. Essentially, it’s a general vocabulary that describes web resources. Dr Kauppinen highlighted the fact that there are no real science entries within Schema.org, which means anyone searching a particular scientific topic might not discover all available information. We broke into three groups to identify how a set of scientific schemas could be built, then came back as a single group to compare results.
What emerged quite quickly was that the question of “can we develop schemas for science?” was secondary to “should we develop schemas for science?” The task of drafting categories for scientific schemas soon gave way to a broader discussion of the need for such a resource.
We all agreed the motivation was justified, and very much in synergy with Dr Kauppinen’s development of Linked Science which aims to connect all scientific resources semantically together. But many issues were raised as to the details of putting the science schemas together. The impetus behind Schema.org is to remain relatively small and simple to enable wide implementation, which would be fine for us if the idea was to make science more accessible to non-specialists. If the audience comprised scientists, however, there would need to be more detail to allow rich semantic searching. And someone raised the important question of whether search engines would even be interested, or whether they’d see science schemas as too specialised.
The workshop certainly provided some good discussion – we overran into the lunch break and nobody noticed for quite a while! The idea of a Schema.org for science is undoubtedly a relevant and ambitious idea, but perhaps the aims need to be more refined before it can be developed into a truly applicable resource.
Bridging the divide: building around the PDF
This workshop could be seen as a natural continuation of Beyond the PDF, which took place in San Diego earlier this year. It was an opportunity to learn about Utopia Documents from the developers themselves: Dr Steve Pettifer and Dr Philip McDermott of the School of Computer Science at the University of Manchester.
Utopia tackles the fundamental problem with PDFs: they are static, locked documents, providing no way to usefully draw out information which a reader may be interested in. Put simple, it’s a PDF reader with bells and whistles. And those bells and whistles are really quite impressive: access to enhanced html content, linking of protein structures to PDB entries, in-page reference linking to PubMed entries, manipulation of data in tables, links to commentaries and the integration of ChemSpider for chemical structures.
This session wasn’t just a promotion slot, though. It raised the issues surrounding the current publication model and the need to provide tools that encourage authors to enhance their content. There were those who strongly felt that PDFs face certain death as the format for viewing papers, those who felt they were necessary; there were strong comments about the strict financial motivations of publishers (which, as an Editor for a society publisher, I must refute as a little harsh) and the language barrier between readers of papers and the computers which are used to access them.
The strongest message that came through for me was that even the most robust paper is merely a representation of a scientific study; it doesn’t show the thought process of the researcher or the wider context of the work. As an example, you only have to look at Watson and Crick’s paper on the structure of DNA, which is possibly the most understated discussion of scientific results ever published! (perhaps link to special archive where paper is available: http://www.nature.com/nature/dna50/archive.html) In order to make the most of current technologies, we have to be able to expand upon what is presented in a paper by enabling the kind of enhancement that Utopia demonstrates.
So many ways to tell a story
It may seem strange to include storytelling in conference about science communication, but to paraphrase Trust-funded researcher Beau Lotto, “The brain deals in narratives…not random facts”. Even scientific papers are a form of storytelling, with a plot and characters, albeit rarely with a firm conclusion. It was this premise that formed the basis of our workshop, facilitated by Anton Zuiker and Bora Zivkovic, both prolific bloggers with day jobs as Communications Director for the Duke University Department of Medicine and Blog Editor at Scientific American, respectively.
Harking back to the panel discussion earlier on the Arsenic Story, there were comments on the problems associated with scientific reporting and the way in which journalistic stories are compiled. There was no real answer, but it became clear that the misrepresentation of science in news stories remains a major concern both for scientists and communicators. Is it the responsibility of researchers to write papers which are more accessible in order to prevent this from occurring? A solution to this is to employ people specifically to write papers, but this in itself raises questions as to what recognition this person should receive.
The topics covered were varied but one key point seemed to tie things together: science is all about making connections. They methods by which we make these connections are broader than ever and perhaps we need to consider how our communication, our scientific storytelling, needs to adapt to support this diversity of audience.
A National Undergraduate Bioscience Research Journal
“Bioscience Horizons” (http://biohorizons.oxfordjournals.org/ or @BiosciHor) is a national journal accepting submissions by undergraduate students and published by Oxford University Press (OUP). In this session, Neil Morris (University of Leeds) and Cathy Kennedy (OUP) introduced the journal and asked the audience for suggestions on how to use social media most efficiently to make it more known amongst students and encourage them to submit their work.
Bioscience Horizon is published online and open access and accepts submissions from all biodisciplines. High quality of the published research is emphasised and, in contrast to other undergraduate journals, articles are rigorously reviewed by UK academics – not by students. Undergraduates will appear as the sole author, of course with support and agreement from their supervisor or academic sponsor. By submitting an article students get first experience with the publishing and peer-review process. If their paper is accepted, it can boost their confidence enormously, as their name is being recognized within the scientific community. It also can help them to make their CV stand out and secure a PhD position.
Even though everybody agreed on the benefits for students being involved in the publishing process, the discussion brought up some concerns. Pre-publication peer-review can take a long time and papers might not be published yet when students are applying for PhD positions. More importantly, what would the incentive for students and supervisors be to publish high-quality undergraduate research in Bioscience Horizon rather than in a mainstream journal? Even though Bioscience Horizon articles have started to receive citations from established journals, “an impact factor still trumps everything else”.
Workshop: Online Communications Tools 2
This was an excellent advanced workshop for those of us using social media widely but often don’t know what tools are best to measure impact (or ‘kudos management’ as one panelist joked).
Brian Kelly of UKOLN at the University of Bath, gave an excellent talk (slides here) analysing how the Russell Group universities are using Twitter and their overall influence. You can see the data and analysis on his earlier blog post.
He used several metric services to compare impact and influence: Klout, Peer Index and Twitterlyzer. Each has some interesting features: Peer Index for example, generates star graphs to show what subjects an account is talking about, and what it’s not. It also provides a table of rankings and these are ‘authority weighted’ — a score of 40+ means you’re in the top 10% of users compared. Lou Woodley from Nature Networks said this is what they use to see if an account is worth keeping open or not. However she emphasised that no tool should be used in isolation.
This was the first hint of scepticism: the talks confirmed to me that most social media metrics are inherently misleading — how to distinguish between accounts that are RSS bots compared to real engaging people? Or larger numbers of followers vs more engaged or influential followers? Microbiology blogger Alan Cann said these aren’t ‘metrics’ but game scores, single number stats. However, he reckons data visualisation could help us gain more meaningful insight into what and how our social media networks are shaping our communications and work. The value is not in the scores themselves, but in comparing them.
In a separate point, Kelly highlighted the value of embeddable content. He showed how embedding a talk slides on Slideshare, thereby allowing them to be shared on other sites freely, meant slides were viewed more than triple the number of times they would be on your own website. They key is making sure you can track those embeds – that’s a useful measure of traffic as much as traditional pageviews.
Kelly also emphasised how speakers are reaching a much wider audience than those they are presenting to in a room, through coverage on Twitter and live streams.
Alan Cann stepped up to talk about how his department has been using Friendfeed to engage their undergraduate students (which reminded me of something Stephen Curry said at a fringe event a couple of nights before – about his department at Imperial using blogs (one for the staff, one for students) to encourage discussion).
Again, Cann emphasises that you must be careful with what you use in terms of summary stats. His analysis, based on data visualisation rather than numbers, showed that female students were more active in online discussions about lectures, but there were no differences in exam marks at the end of the year. He posited that males focused on the assessment part of learning – they do just what is necessary and no more. However, the greater engagement of their female peers may mean they get more educational benefit from the system.
Cann also highlighted the difference between ‘visitors’ (those who visit once a day at the same time) and ‘residents’ (those who interact) — something to bear in mind for websites, blogs and Facebook pages.
Speaking of Facebook, one discussion point was why scientists don’t use the world’s biggest social networking tool for collaboration and discussions. A good point was that for scientific collaboration is that it simply doesn’t work. The best model is to take a sound platform and look at options for plug-ins to facilitate what scientists want to do — this is possibly why Google Wave was so exciting for scientists (who saw it as, among other things, a way to write papers collaboratively across the world and with the facility to write bots that will pull in data, citations, do calculations etc. automatically).
Additional reporting by Dr Vibhuti Patel and Anne Osterider.
Want more? Full list of blog coverage of #Solo11 is available on their wiki.
Researchers: where are you having online discussions about your work? We’d like your input on this post.