 |
|
| FriendFeed is the easiest way to share online. Learn more » |
#blts please don't leave Orlando! - http://twitter.com/bratmed...
#blts not @blts - http://twitter.com/jrambo0...
I can't find Kevin Hazlett on youtube. Anyone? #blts - http://twitter.com/HowardS...
Enjoying some #blts and the yanks game, packaging up those fast-selling twitter/tweezer shirts #phish - http://twitter.com/phanart...
#blts More (delayed) from John Wilbanks: "Tweaking for the end user is probably revolutionary for the back end" - of scientific publishing. - http://twitter.com/peterba...
#blts More (delayed) from John Wilbanks: Scientists sharing their 'craft', as opposed to results, is usually not a problem. So can be open? - http://twitter.com/peterba...
At last minute went to see John Wilbanks at BL #blts. Not especially inspiring just, er, right, and good responses in discussion after. - http://twitter.com/stevehi...
#blts Wilbanks: "Publishers become code libraries". Interesting idea. Clearing the way for open access? Sounds double-edged. #oa #openaccess - http://twitter.com/stevehi...
#blts Wilbanks: "coding of science is important", in response to @petermurrayrust question: why fear getting open science wrong? - http://twitter.com/stevehi...
@petermurrayrust data, along with grant proposals, are amongst the most hidden aspects of science - making them open will change much #blts
Deepak Singh, Bill Anderson and Jean-Claude Bradley liked this
RT @CameronNeylon Scientific Findings in a digital world: John Wilbanks at the British Library http://ff.im/5ziee #BLTS #eresearch - http://twitter.com/jimrhiz...
Following British Library seminar "Scientific findings in a digital world: What is the genuine article?" w/ John Wilbanks: hashtag #blts - http://twitter.com/JMarkOc...
Lots of good stuff at #blts but hard to see where change will come from while funding rests on trad publishing - http://twitter.com/lzsl...
RT @steinsky are we all agreed: elseveir 'article of the future': just an ugly website redesign? #BLTs - http://twitter.com/boudicc...
but that's all coke is: a brand people trust cos of cute santa ads even though it's just worthless sugar water #BLTS - http://twitter.com/steinsk...
@petermurrayrust data, along with grant proposals, are amongst the most hidden aspects of science - making them open will change much #blts - http://twitter.com/EvoMRI...
A recent blog post by Peter Binfield on printable impact measures: http://bit.ly/BvF8q #BLTS - http://twitter.com/EvoMRI...
RT @dbkell: #blts the problem is that most folk do not read (let alone cite) the literature - see e.g. http://bit.ly/OVtjm - http://twitter.com/EvoMRI...
"Scientific Findings in a digital world: What is the genuine article?" John WIlbanks at the British Library - liveblogged here #blts
Michael Habib, Bora Zivkovic, Rafael Sidi and 10 other people liked this
So we are all in and the speaker is here. Interesting set of people in the room. Publishing, funding, policy, trouble makers...
- Cameron Neylon
liveblog commences
- Graham Steel
there is a big screen up the front showing the BL room - I'd forgotten that the default UI would show a big smiley picture of me if I set up the item :-)
- Cameron Neylon
Roughly how many folks are there, Cameron?
- Graham Steel
There is probably about 60-70 people in the room I would guess
- Cameron Neylon
Just starting now...introduction to the BL
- Cameron Neylon
Describing digital initiatives at the BL - Shoutout to UK PubMedCentral, beta version of the website coming soon
- Cameron Neylon
BL working with MSR on virtual research environments #blts
- Fiona Bradley
Twitter using #blts will be automatically streamed into the friendfeed room
- Cameron Neylon
Sarah K introducing John - thinking for a while about having a session on the future of scholarly communication
- Cameron Neylon
Cameron - yes please!
- Fiona Bradley
....and heres John!
- Cameron Neylon
Going to be a cut through what is a big topic
- Cameron Neylon
Wilbanks: huge topic, so will do his best Oprah Winfrey in the discussion
- Fiona Bradley
What is scientific "knowledge" - showing a journal page from late 19th C
- Cameron Neylon
Communicated knowledge by writing it down, printed on paper, and posted out to people
- Cameron Neylon
Journals haven't changed that much over the years, only the technology we use. Format is the same
- Fiona Bradley
Showing a journal from Hindawi - really not very much has changed, incremental advances over the printed page article
- Cameron Neylon
The earhorn example as an example of the pitfalls of incrementalism - see one of John's online talks at http://slideshare.net/wilbanks
- Cameron Neylon
'incrementalism' paper to PDF.
- Fiona Bradley
Overcoming incrementalism is a real challenge - large shifts require big changes in thinking. There is much more to scientific knowledge than the printed paper
- Cameron Neylon
research materials, data, ontologies, annotations, wikis
- Cameron Neylon
What do we do with ontologies, annotations and wikis? Semantically enhanced articles are one way (yes!)
- Fiona Bradley
showing a semantically marked up paper - the one from PLoS Neg Trop Dis that was described in PLoS Comp Biol
- Cameron Neylon
Google works because hyperlinks are sparse but with a fully marked up paper there are very dense links. May need to do something different, as well as "just" doing the markup.
- Cameron Neylon
Sem articles can also include raw data from that article, and relevant data from other work.
- Fiona Bradley
Now showing the http://beta.cell.com - it looks new. "It reminds me of when AOL told us what the future of the web was and it looked a lot like TV..." Want raw data, real information.
- Cameron Neylon
We need radar, not earhorns, it is going to look very different and be disruptive. Just like radar was.
- Cameron Neylon
Elsevier's Article of the Future project. This reminds me a bit of Herbert Van De Sompel's work of a few years ago on linking from data through to article and subsequent citations.
- Fiona Bradley
jove - peer reviewed instructional videos. Web is "forcing its way into science"
- Fiona Bradley
scivee is a OA alternative
- Fiona Bradley
If you'd said in 1992 that the web would destroy bookstores you wouldn't have been believed. Don't believe people when they say they know what the future looks like.
- Cameron Neylon
Four princples of publishing: registration, certification, disseminatoon, preservation. Regisstration and dissemination obviously broken by web
- Cameron Neylon
Now showing my lab notebook. Woot!
- Cameron Neylon
Woot
- Graham Steel
Multiple copies of web data makes it harder to fake/change data.
- Fiona Bradley
Instant registration and dissemination. But that is not enough. Otherwise could set up a random statement generator and wait for the Nobel prize to roll in.
- Cameron Neylon
Preservation: old way was the library. But we don't do so well online. LOCKSS is a good initiative.
- Fiona Bradley
Preservation in the digital world is not so good. Internet archive is not comprehensive. Mention of LOCKSS program. Multiple copies with hashes.
- Cameron Neylon
Peer review combines two things and confuses them. 1) Is it sound? 2) How important is it?
- Cameron Neylon
First is (hopefully) more objective than the second. Showing a funny slide out peer review. I don't know the source. "Strange Matter"?
- Cameron Neylon
is a piece of work impactful? Very complex. Peer review is there so we are not disproven?
- Fiona Bradley
new work displaces knowledge - it is harder to get the first paper published on a topic than the 3rd or 4th.
- Fiona Bradley
PLoS ONE example - separating validity from impact - can anyone point me to the paper that correlated a persons publication impact with the likelihood of rejecting papers? Just got queried but can't find in Friendfeed right at moment
- Cameron Neylon
showing the sem web stack, and discussing the legal aspects of the web. "what is a copy?"
- Fiona Bradley
Power of standardisation: TCP/IP and HTML examples. How to do for scientific information? Need three layers, technical, legal, and semantic. Semantic web doesn't really work yet. Legal infrastructure is getting there but issues as science is often not about making a copy - therefore copyright doesn't apply
- Cameron Neylon
FYI, if following on Twitter, petermurrayrust and peterballantyne are both live tweeting the event. Use #BLTS at search.twitter.com.
- Jill O'Neill
Annotation - a semantic problem. Giving the coffee exampls: again see http://slideshare.net/wilbanks for this slide sequence
- Cameron Neylon
There are as many names for genes as there are names for coffee
- Cameron Neylon
Need common names to be able to share anything semantically.
- Cameron Neylon
need to agree on common names for things to create an ontology. common annotation needs common names. last challenge is federation.
- Fiona Bradley
Federation: There are at least 1000 databases with over 250 different terms of use for nucleic acids. BUt using climate change/polar info from Polar year project. Lots and lots of pieces. Trying to make informed decisions about climate change. 399 different databases with different terms on reindeer herding and climate change?!?! Did I really get that right?!?
- Cameron Neylon
Discipline exceptionalism: "No one else appreciates how complicated our work is, therefore we need to write our own definitions and terms and policies"
- Cameron Neylon
better to not do things, rather than do them wrong, with disastrous consequences?
- Fiona Bradley
If we get it wrong - won't just be garbage in garbage out, but amplified garbage out. Systems could be unstable. Single copyright case brings the whole thing down. Wrong protein to wrong gene, people get sick and die.
- Cameron Neylon
legal, techical and semantic resistance points in the wire before we have genuine articles coming out of the system.
- Fiona Bradley
Moglen's analogy to barriers to open source. Ask not why the electrons want to move around the circuit. Ask what the resistance of the wire is? Key Question: What's the resistance and how do we overcome it?
- Cameron Neylon
...applause...
- Cameron Neylon
what is the resistance to open science, and how do we overcome it? technical solutions alone won't solve all issues
- Fiona Bradley
discussion coming up in a bit
- Fiona Bradley
anybody out there wants to ask questions then drop them in here...
- Cameron Neylon
Thanks Cameron & Fiona
- Graham Steel
you're welcome graham, send you my chiropractor's bill later :-)
- Cameron Neylon
wondering how many other people here are library people (other than BL staff of course!)? what can we do to help develop open science?
- Fiona Bradley
Hmmm, not sure how many are library people per se. Most seem to be publishing or academic trouble makers...
- Cameron Neylon
gathering up the crowd for questions
- Fiona Bradley
today's OKFN virtual meeting is now in full swing - can I keep up both? I'll try :)
- Graham Steel
Q: Author ID. How are people associated with the work they do? (huge issue!) A: Open ID unsatisfactory. there are several other services. Libs and publishers the right providers for this (note: some libs working on name registers)
- Fiona Bradley
honesty a huge issue in science. This issue has to be resolved. when people generate data, they should create carefully using common names for things from the start so data can be interoperable.
- Fiona Bradley
John: Its a complex problem, solution probably going to come from a stable institution, Universities, governments, publishers. (Me: Signfiicant question mark over universities' stability though)
- Cameron Neylon
John: Deciding what to throw away may become the big problem as data explodes
- Cameron Neylon
Who will store massively huge data sets? When do you make the decision to delete?
- Fiona Bradley
what made the web was good timing, and openness.
- Fiona Bradley
Complicated question about: "to what extent can you draw an analogy between 'mixing up scientists and the web and seeing what comes out free' and natural evolution". John answers talking about walled garded net services circa 1992. The web won out, in an evolutionary sense, because it was open and anyone could play.
- Cameron Neylon
lots of dumb ideas will die, lots of good ideas will die bcause of bad timing, but some will survive. the vision for semantic articles now is unlikely to be how the eventually look
- Fiona Bradley
Q: researchers aren't really predisposed to share? A: but publication by patenting is disclosure and sharing.Issue of acclaim and notice to the author. This gives the incentive to share. Need to be able to track impact of blogging, sharing data and connection with impact and acclaim. Then they will share. Need new ways to get credit in science for your work
- Fiona Bradley
Stephane Goldstein: "Is there a mistaken heroic assumption that scientists like to share" [giggles from audience]. John: Publication and patenting are sharing. Not because people want to but because people get registration and validation. Therefore a question of providing incentives to share materials and links. Key is enabling people to claim credit through things other than...
more...
- Cameron Neylon
Q: in this interconnected world, is the structure of academic articles stlll valid? A: yes. procedure, method, abstract still valid.(lots of raised hands for extra comment!)
- Fiona Bradley
comment from audience - data needs to be truly available because it is generally not possible to recreate someone else's work and test their hypothesis without cost.
- Fiona Bradley
reading through the thread... "a session on the future of scholarly communication" sounds like a great topic for future shindings
- Daniel Mietchen
comment from audience: no one really reproduces experiments though. they work to test hypotheses of others. John: but the parts you need to reproduce are the things you need to build. Research needs to be reproduceable (and open?) - even if not done in practice
- Fiona Bradley
papers as an advertisement for research? (discussing comments by Alma Swan on future of article - the new role of science blogs etc)
- Fiona Bradley
Q. from the OKFN session that I would like to put to John and post back "Can we tell a story about people who have been helped by open data in development yet, or is it still theoretical"
- Graham Steel
batteries about o go here. Ok - will try to ask that question - schedule is slipping here. John doesn't like making imprecise short answers...
- Cameron Neylon
comment:not all data is bad data. But need to be able to spot systematic errors. Data needs to be annotated.
- Fiona Bradley
q standards for dscribing articles. eg mods, mets. who is going to fix it? A: look at screeds against HTML in the early days. The simplest standard that is extensible and correctable wins. But there is also a certain amount of luck involved.
- Fiona Bradley
sem web will allow us to map contributors in one piece of data with authors in another: find both. But must be able to weed out what is merely similar.
- Fiona Bradley
q: peer review should evaluate impact and quality together? a: this is the best reason to do peer review, so you know what to trust like brands - like you might trust coca cola.
- Fiona Bradley
"being a peer reviewer is a difficult burden, being on an editorial board looks good on your cv"
- Fiona Bradley
and what if, instead of listing editorial boards on your cv, everyone could see the reviews you did in public?
- Daniel Mietchen
q; key motivation for researchers is street cred from their peers- right now that is through citations but there are other models. how do you build these new mechanisms in a networked world? a: something that can be printed for a review meeting that shows impact?
- Fiona Bradley
q: do we have a responsibilty to make articles accessible in the online environment?
- Fiona Bradley
comment: still a need to constrain the size of publications so that they still make their point, but also using the potential of new technology well
- Fiona Bradley
q: role for the BL in this? a: work on DOIs for data, unique names for things is essential. norm setting role for libraries as a partner with publishers.
- Fiona Bradley
comment: this forum provides a neutral ground to discuss these issues, is good. (valid point!)
- Fiona Bradley
q: there is too much to read! traditional method of reading needs tweaking. if your lib doesn't subscribe, you can't access materials: a: open access is the solution (laughter). w use journal brand as a proxy to help determine what to read. we need to be able to evaluate information is it worth reading.
- Fiona Bradley
great talk thanks to John, the room and the BL
- Fiona Bradley
constraining size of publications is certainly useful, but could be much more effective if we'd replace the introduction of manuscripts by hyperlinks to collaboratively written encyclopedic articles that are updated as new results come in - http://bit.ly/4ftzn . Similarly, the methods and results sections could simply link to the relevant data pages at places like OpenWetWare, and the article itself could focus on the discussion aspect.
- Daniel Mietchen
petermurrayrust: #blts interestingly much of the discussion is about data, not articles - http://twitter.com/petermu...
petermurrayrust: #blts interestingly much of the discussion is about data, not articles
- Richard Akerman
More microblogging from this TalkScience event: see friendfeed.com/British-library #BLTS - http://twitter.com/brunell...
#blts interestingly much of the discussion is about data, not articles - http://twitter.com/petermu...
Incremental tweeting the exception? RT @brunella We're in a disfunctional info age - with all this incrementalism in communication #BLTS - http://twitter.com/peterba...
Other ways to read this feed:
