Science in the Open - Latest Comments

Re: The Nature of Science Blog Networks

CameronNeylon — Mon, 26 Jul 2010 05:58:55 -0000

Absolutely! Coffee and in person are good as well!

Re: The Nature of Science Blog Networks

Eva — Mon, 26 Jul 2010 05:49:06 -0000

Great post. Insightful as usual. Instead of leaving a comment, can I comment in person at SciFoo over coffee? I know you like everything to be online and accessible, but I like coffee and meeting in person. Ha!

Re: Capturing and connecting research objects: A pitch for @sciencehackday

Janne — Fri, 09 Jul 2010 10:55:48 -0000

The problem isn't mainly how to share data. The basic problem is that there's few to no incentives for sharing your data in the first place, and many strong incentives against.

Re: Metrics of use: How to align researcher incentives with outcomes

CameronNeylon — Thu, 24 Jun 2010 14:27:56 -0000

Danny, you won't get any argument from me on that. I think evangelization has a role because it helps to raise the profile of the issues but on its own it is not enough. Granting agencies have to start actually caring about optimising output in the general sense - and my feeling is that that will require as you say looking at a much wider range of outputs.

Re: Metrics of use: How to align researcher incentives with outcomes

Danny Navarro — Thu, 24 Jun 2010 12:59:17 -0000

Granting agencies have to reward more data generators. It seems all the glory goes to the people who use the data to publish (who in many cases don't credit the data generators properly).

In my field, proteomics, sharing data is not only more work for the researcher but it also gives advantage to a competitor who might get it out before you and will get the next grant in the field instead of you.

I don't think evangelization will work. Granting agencies should start looking into something esle that is not publications per se.

For example, nobody cares about making programs maintainable or fast because you get the same kind of publication with a crappy program than with a good one, as long as both have the same function.

Re: Metrics of use: How to align researcher incentives with outcomes

Marius Kempe — Mon, 21 Jun 2010 05:52:11 -0000

Thought you might be interested in this short article from evolutionary biology, along very similar lines to what you've written.

http://www.nature.com/nature/journal/v441/n7093...

Re: Capturing and connecting research objects: A pitch for @sciencehackday

danw — Wed, 16 Jun 2010 11:12:59 -0000

Could you use Pachube? http://www.pachube.com/

Re: Metrics of use: How to align researcher incentives with outcomes

Mk630 — Fri, 11 Jun 2010 05:21:23 -0000

Very nice post. Thought you might interested in this, in case you hadn't seen it yet: http://www.biomedcentral.com/1756-0500/2/113

Re: Implementing the “Publication as Aggregation”

facebook-818642519 — Tue, 25 May 2010 12:27:34 -0000

Hi Cameron, great blog! The ideas you propose in your blog are similar as I have seen being developed in the Netherlands by various institutions and supported by SURFfoundation. The joint effort of several Dutch universities have created a lot of knowledge and tools for creating so called "enhanced publications" which use OAI-ORE to describe the semantic relations between scientific web-resources.
read more: http://bit.ly/au3HkR

Yes, There are author tools for these aggregations! Many are domain specific, but one tool is generic that came out of the tender projects. see: http://escape.utwente.nl/ It is public so try and test yourself. If things go well there might be pilots running for creating aggregations during the publication process for open access journals, transferring the supplementary material to long term preservation archives and preserving the link interity by making use of persistent identifiers.

Work that needs to be done:
Regarding visualisations; we distinguish two kind of visualisations. 1. the visualisation of the data (like IBM's Many-Eyes, or Google's Gapminder) and 2. the visualisation of the aggregation/the relations interconnecting scientific web-resources. These visualisations could possibly be generically available in the near future using Rich Internet Applications, in order to embed in webpages. We are preparing to work in cooperation with several universities and institutions for semantic web and media design.

Regarding OAI-ORE Datamodeling: in order to create visualisations for generic use, like HTML, the RDF must be modelled according to specific rules. Also here several universities are working together to create a more strict standard, at least for generic use by the Dutch universities. Apart from these core set of rules there is room for extensions that describe the specific domain-related stuff. This then can easily be added in the RDF and used by domain specific visualisations.

Citing: for citing trusted scientific content that scales, millions of persistent identifiers can be used in the urn:nbn namespace, for publications, dataset, or even cells in a table. These id’s are minted by the national libraries who create a trusted information space, guaranteeing technical quality and persistence of the content. Using resolvers, persistent identifiers and a strongly committed preservation policy, the content can be independent for future (re-)location, duplication and future technology. These Persistent Identifiers make it ideal for Enhanced Publications to refer to scientific webcontent, creating trust and reassurance in future web-presence and availability of the digital content.

Link-back citation: scientific web-resources are themselves unaware they are used as part of an aggregation. The aggregation, created in an authoring or storage environment, should notify the resource that is has been used as part of the aggregation. This mechanism can be automated using the XML-RPC protocol for Ping-back, (less spam sensitive then track-back) which is also used in modern blog systems. The web-resource keeps track of the aggregations it is part of creating a ORE resource-map itself, with all the URI’s of the aggregations it is has been told to be a part of. Making use of ping-back, automatic citation discovery on resource level will occur.

Please do let me know what you think of these developments.

Re: Implementing the “Publication as Aggregation”

CameronNeylon — Mon, 24 May 2010 16:15:10 -0000

I guess I struggled a bit with Utopia because it is in the context of a pdf which just seems completely the wrong direction to me. Seems to me that display isn't a problem really, the web can do it all already, and I didn't really understand why they didn't expose all of that information more on the html version. For me it's the authoring side that needs the work. On the other hand the continued preference for PDF does seem to be pushing things in the direction of laid out content. See under iPads and magazines for instance.

Re: Implementing the “Publication as Aggregation”

openid-23918 — Mon, 24 May 2010 12:09:43 -0000

Have you looked at Utopia Documents for an idea of how this could look. They offer a lot of this kind of functionality in the here and now.

Re: The future of research communication is aggregation

CameronNeylon — Mon, 10 May 2010 03:41:04 -0000

I disagree on two counts here. Firstly I never meant to say a paper is
simply a collection of information. In fact I would argue that my intention
is to raise the status of the composition process to primacy so as to enable
more creativity.

Secondly the role of the artist is arguably in choosing the _right_ set of
constraints for a particular composition. Constraints can be powerful and
focussing or they can be unhelpful or dangerous. You assume that our current
straighjacket is the right one. I would argue it is more like playing tennis
with one hand tied behind your back. The majority of papers published today
are _less_ than the sum of their parts. Ideas, data, and imagery that are
damaged, made less clear by being tied into an old fashioned framework. You
are saying that all poetry should be sonnets. I am saying that at the moment
we have haiku being rammed in as the heroic couple because sonnets are the
only way someone can get published

But I think you're iPhone analogy is apt. The iPhone is a closed system that
focuses on producing slick user experiences that match their expectations.
What it is very bad at is supporting radical innovation because the whole
environment makes taking risks and exploring new ideas difficult. It
channels behaviour and thinking in defined ways. This is exactly the
opposite of what we want in an optimised scientific environment. In science
we need generative systems that support the process of innovation, while
efficiently identifying blind alleys and mistakes. Science needs the madness
of open source. Technology and marketing can benefit from slick interfaces
and the iPhone approach. That is engineering, not science - and they have
very different needs.

Re: The future of research communication is aggregation

JohnMorris — Sun, 09 May 2010 17:27:11 -0000

This is so wrong. Thinking of a paper as simply a collection of information is like thinking of a poem as simply a collection of words. Working within the constraints, of metric, rhyme and so on, struggling with the constraints, you often end up with a more beautiful poem. The same is true of any kind of authorship, with poetry just an extreme example.

Actually, it holds even more broadly. Compare developing on an iPhone to developing for a Blackberry, for example. Apply loosely enforces a set of interface constraints, and this makes life easier for both the developers and the users.

Re: In defence of author-pays business models

CameronNeylon — Tue, 04 May 2010 16:09:51 -0000

Hi Iddo, and thanks for the comment. I think you've actually cut to the core
of my argument. The point is that actually all of these academics _are_
paying for those subscription access - they just choose to ignore it. And
they're not going to be able to ignore it for much longer. In the next year
librarians are going to be cutting subscriptions back savagely and the
option will be simple: If you want JACS/Biophys J/Cell/J Neurosci then the
department will have to find the money directly. Once that happens my
suspicion is that OA with APCs starts to look attractive.

The catch is that politically the appeal here is that the academics can claw
back money out of the hands of central administration (i.e. the library).
This is an easy story to sell within departments but a longer view I think
includes the realisation that we need the library to manage the transition.
But believe me, the next 18 months are going to look very interesting. There
are many major universities in Europe that do not subscribe to Cell. My
institution no longer gets Science. I can't read most of the journals core
to my work with our institutional subscriptions. It's going to spread and
it's going to start cutting into the core subscriptions of research
departments who are then going to have to choose whether to put the money up
themselves. That's where change starts, when the pain bites.

Re: In defence of author-pays business models

Iddo — Tue, 04 May 2010 15:18:25 -0000

Preface: I am a great supporter of open access, and I agree that most arguments against that model are bunk. Given a choice, I would pay more and publish in an OA publication. At the end of the day, though, whatever hidden costs their may be do not directly manifest themselves to the author, closed-access is still cheaper than open access for immediate out-of-grant-pocket expenses. Page charges nonwithstanding. If one is fortunate enough to have an extramural grant that is budgeted for OA publications, then that works fine. If not (and many do not, because small size grants don't have enough bandwidth for the expected $2500/yr or there is no grant and you are running on institutional funds, as happens in many R2 universities) most groups would rather spend this money on research and/or student stipends.

Yes, most journals would waive the fee if the author claims hardship. But most authors would not claim hardship: they have the $2500, they would rather give it to a summer student. So they would go with an equally reputable CA publisher, maybe pay $500 in page charges.

The cost-benefit of open access is still unavailable to most people, but the OA advocates do not seem to realize it. Unless you have have an R01 with OA charges in it, you have no compelling reason to go with OA. Ideological support for a laudable cause nonwithstanding. Yes, the ideas are great. No, the immediate benefit is not apparent. Maybe the library overhead does may CA more expensive in hidden costs than it is. So? It is still there!

Point being: most of my colleagues would not publish in an OA journal due to the cost. In fact, all they know of OA journals is that they are expensive. OA seems like a fringe benefit, especially in the current promotion and tenure criteria in universities.

Bottom line: there is between OA advocates and a large part of academia (which publishes CA), analogous to the one that exists between open source advocates and the rest of the world (which uses Windows). There is a fundamental misunderstanding of what OA is about, and the lack of ability to convince people to pay for it when CA is perceived to be cheaper.

Re: In defence of author-pays business models

twitter-18504532 — Sat, 01 May 2010 09:04:36 -0000

Thanks to Cameron Neylon for his useful references to estimates of the percentage of rejected papers that eventually get published in another journal (50% - 95% depending on field). And of course there's always the fallback of unrefereed book chapters!

[Lest there be any misunderstanding, my own remarks about refereeing standards and pay-to-publish journals("Symptoms of Premature Gold OA -- and their Cure" http://bit.ly/b8vS1a ) are generic, not specific to PLOS journals (in which I myself publish!).]

See: Beall, Jeffrey (2010). "“Predatory” Open-Access Scholarly Publishers". The Charleston Advisor 11 (4): pp. 10-17(8). (Alas not OA!) http://bit.ly/9dKvk8

Re: In defence of author-pays business models

openid-16041 — Sat, 01 May 2010 05:35:20 -0000

Hi Cameron,

"That is why, within reputable publishers, structures are put in place to reduce that risk as far as is possible, divorcing the financial side from editorial decision making, creating Chinese walls between editorial and financial staff within the publisher."

In an ideal world, this whole issue is irrelevant: in an ideal world, the content of the paper would be reproducible, and it would not matter if it was sponsored or not; anyone would have no problem on deciding on the quality of the paper. Any structure put in place by publishers is nothing more than a dirty hack working around the real problem that scientists do not generally care about reproducibility.

Re: In defence of author-pays business models

Peter Binfield — Fri, 30 Apr 2010 11:37:18 -0000

Cameron - Completely correct. Only our finance staff see that info

Re: In defence of author-pays business models

CameronNeylon — Fri, 30 Apr 2010 04:28:17 -0000

Thanks Ramy. I also note that your comment on the post itself makes many of
the same points. I didn't see your comment until I'd written my post but
they're worth reading together I think.

Re: In defence of author-pays business models

CameronNeylon — Fri, 30 Apr 2010 04:27:36 -0000

Indeed I believe the professional editorial staff don't even know whether a
waiver is requested or granted. Again, you set up management arrangements so
that any potential conflict of interest is avoided.

Re: In defence of author-pays business models

Anna Croft — Thu, 29 Apr 2010 18:24:53 -0000

Regarding editorial decisions and whether the author pays or not, it should also be noted that in fact the AE doesn't even know if an individual author pays - as the author may have requested a fee waiver.

Re: In defence of author-pays business models

openid-23163 — Thu, 29 Apr 2010 17:09:32 -0000

Excellent.
There is not a word in this post that I don't agree with, notably those about the role of AEs in PLoS ONE, and the unjustified attack on their integrity and expertise.
Thanks Cameron

Re: The future of research communication is aggregation

CameronNeylon — Thu, 29 Apr 2010 14:30:37 -0000

Sorry, I hadn't twigged what you were talking about. The "discussion section" of a traditional paper isn't there simply because I forgot to put it in! Yes, I take your point that this view enables a much more flexible way of imagining this (and indeed the other sections - methods might be an aggregation of code and scripts, or a makefile). Yep a dynamic diagram would be cool - a bit beyond my abilities at the moment though...

Re: The future of research communication is aggregation

msscha — Thu, 29 Apr 2010 08:56:38 -0000

As a reader of science, I like the idea of a "curated discussion" but I also
want space explicitly provided where the researchers themselves have the
opportunity to say what they think their research accomplished or how it
fits into the bigger picture. As writers of science, my students are often
dismayed when they get to the Discussion as a writing object b/c it is much
more limited in scope than they were hoping -- I think this idea of a
dynamic paper addresses this problem nicely b/c it "detaches" each part of
the paper, meaning that a discussion can exist with criteria attached AND
there is still room for the looser, more interesting speculation that any
piece of research also provokes (the stuff I tell my students happens in
hallways and at conferences, especially at the bar). Also, just want to say
that I really like the diagram -- it's taken most of the last year for me to
finally "get" this approach. Would love to see the figure animated, a la
the kind of stuff Thinkmap puts together for the visual thesaurus and other
data-heavy projects (www.thinkmap.com).

Re: The future of research communication is aggregation

CameronNeylon — Wed, 28 Apr 2010 15:01:22 -0000

Hi Mickey

Its a good question. I think discussion is actually implicit in the diagram.
The overall discussion is the set of all other communications that cite some
portion of the original. You can then imagine reducing that large set by
specifying all communications that come from some subset of sources
(friendfeed and twitter, or formally published reviews), or that cite just
the conclusion (or just the data, or...). By following the citation trail
you can also actually get a threaded discussion (or a webbed one if there
are cross citations).

An editor might aggregate some subset of these together to create a new
communication which would be a curated discussion based on some criteria as
well. But all in all this kind of view does some nice things like explicitly
connect formally published and informally published objects together.