Because the performance of this open blog was suffering under FriendFeed’s frequent outages, I was forced to remove included bits from FF from the sidebar (they were for ’scspn’ and ‘openscience’).
I tried to save them, scraping via RSS and other means, but the generally poor performance at FriendFeed was in the way of that, too. I’m sorry to see FF go, in more ways than one. BG
With many thanks to Lee Dirks @ldirks of Microsoft External Research and Lisa Green @boudicca of Science Commons, we learnt today that videos and full presentations of February 2010′s Science Commons Symposium are online. By full presentations, I mean close-up views of every slide, better in that regard than even having been there personally.
Their quality is so good that I plan to revisit every talk. I may retouch earlier notes and add more meta.
A reminder: this is an open blog which means you are welcome to comment, publish posts, edit posts, add links, and so on. All versions of all posts are saved forever, like a wiki, so have at it. On the right under “Join In” I have updated the required steps.
Quoting from Lisa’s emailed announcement:
Watch them here:
Session 1 http://content.digitalwell.washington.edu/msr/external_release_talks_12_05_2005/18174/player.htm
Session 2 http://content.digitalwell.washington.edu/msr/external_release_talks_12_05_2005/18175/player.htm
Session 3 http://content.digitalwell.washington.edu/msr/external_release_talks_12_05_2005/18176/player.htm
Session 4 http://content.digitalwell.washington.edu/msr/external_release_talks_12_05_2005/18177/player.htm
The videos will play in any browser, but please note that you will need Silverlight and Windows or OSX for them to play properly.
Videos may be downloaded as .wmv here:
Session 1: http://bit.ly/9NJUoL
Session 2: http://bit.ly/bwuntB
Session 3: http://bit.ly/9QIt1G
Session 4: http://bit.ly/aDKv7i
John Wilbanks’ slides for this presentation are at:
http://www.slideshare.net/wilbanks/seattle-sc-symposium-2010
He stresses that many of the speakers today share slides, material, ideas freely.
Creative Commons and Science Commons have existed now for 5 years. Why did cc begin? “Consumers do more than consume.” Making is regular now, but was radical then. The large majority of creative work was illegal under the old copyright law. cc began with the intention of legalizing creation.
They began by building a database of creative commons licensed objects; that became untenable, and they realized at one point that “the web was the database.” At that point, use of the licenses took off and now, they estimate exceeds one billion objects. The way they wrote the licenses has exceeded national boundaries and all expectations.
In science, there was not the same criminalization problem — science has always been valuable when given away and by being given away. It has been conservative and effectively pre-technical and needs
“We want Wikipedia for science,” Creative Commons heard, but they also wanted more, like a generic platform for innovation in science, a reframing of the Internet itself, for science.
The goal is to spark generative science. Wilbanks says “open” and “free” are loaded. He cites Jonathan Zittrain “Genrativity is a system’ss capacity to produce unanticipated change through unfiltered contributions from broad and varied audiences.” Failure is expensive in science and so science has resisted generativity, Wilbanks says.
(I would say there are many more, and more powerful reasons why science is internally conservative than the cost of failure. But then I also think “generative” is a less effective description of these same ideas than “open” : )
For a system to be generative also requires being:
The cost of collaboration and of failure is so low in technology, that a few guys could start Twitter in a week or two, with a simple but good idea. Creative Commons wants to bring those low costs to science.
If we want generativity, we must deal with property rights. We cannot ignore the law.
Where the law intersects with science:
data –> secrecy, sui generis
tools –> contract, patent
narratives –> copyright
“Open” and “free” worked in copyright but do not work as well in data and access, being “open” and “free” can break the commons (how is not indicated at this point).
The most important argument for Open Access? http://www.soros.org/openaccess/read.shtml
Creative Commons licenses for data? At a minimum we could do attribution, right? but one article from Wikipedia has 27 pages of attribution now, imagine that problem multiplied by all Wikipedia articles and time. This could cripple innovation.
Attribution cannot scale, but citation can if we only cite those to whom we owe the most.
No Copyright: CC0 1.0 Universal is the only workable answer. It could be a leap, that far from the security blanket. The life sciences community has taken this well, however, as with
The Open Knowledge Foundation actually makes data licenses, which CC are not big fans of, but the two sides still agreed on the Panton Principles.
Generativity requires not only dealing with data and licenses, but also deal with tools and inventions.
Creative Commons set out to build tools to deal with biological materials, including modular concepts like no clinical use, and simplifying it down to iconography and legal code. There is no intellectual property, but there is physical property and it, too must be dealt with when publicly funded and a part of the commons.
Success:
100,000 lines of stem cells are now under CC,
mapped to 100,000 genomes under CC,
mapped to 100,000 histories under CC
The Creative Commons patent project: reconstruction of the research exemption model patent license. This used to be the law, but the courts took it away, so we need to reconstruct the research exemption. A few major companies have committed their research rights but it’s still not a model patent license. With this patent, we need to allow a revenue stream (revenue clause) and allow limitations and exceptions in the field of use. Nike could allow their patents to be available, but only outside of their markets (shoes for example) which would enable unexpected and noncompetitive innovation.
Finally, for generativity in science we have to deal with infrastructure. “We used to produce data faster than humans could structure it. Now we produce data faster than computers can structure it.” — Bruce Sterling (I strongly disagree; I’ve seen some great work in structuring and dealing with Big Data and the night is too young to declare defeat.) We can only hope for one of two paths, Wilbanks says:
… and only the first path is realistic, and even the end of the second path is more likely to be reached through the first path.
in RDF, law is code and code is law
ccREL — Creative Commons Rights Expression Language
John’s larger point: we cannot deal with just one part of this problem, we need to step through all 5 layers and deal with it systematically. To be generative we must be all five of: leveraged, adaptable, accessible, simple, and transferable. He cites the failure of the International HapMap Project, inspired by GPL and “free” software to address this type of problem.
DIY Biology — that science is in the earliest days of its democratization. DIY Genetic Engineering is here, too with standard biological parts which you can download freely.
Is biology about to undergo computer science’s revolution of the 1980s? Will it be an iPhone or a PC? Something safe and entertaining but locked, disabled or something we can all make our own and with which we can change the world and ourselves, if it is less attractive?
Generativity offers us an innovation-based chance of success.
http://www.slideshare.net/wilbanks/seattle',description:'John Wilbanks’ slides for this presentation are at: http://www.slideshare.net/wilbanks/seattle'})">
Peter Binfield from the Public Library of Science, offers in introduction:
Their business is rethinking the academic journal, not much changed since the 17th century from whence it came.
See the slides of this presentation at: http://tiny.cc/ALM12
The traditional functions of a journal have been:
journals more modernly, also:
Only certification (peer review) and filtering for quality are difficult, from those six functions above. Of peer review: only a handful, maybe only 2 participate. In filtering for quality, authors generally waste time submitting papers for review and rejection by progressively less prestigious journals. An example paper considered at http://tiny.cc/ALM10 took nearly 4 years to get published, after multiple rejections beginning with Nature.
Why should a paper wait so long during which time it could have benefitted science and society?
What is the answer? PLoS ONE of course : )
Open Access with the widest possible dissemination
Online only — with no size, topic, nor scope limitations
A scalable business model includes a relatively minor ‘publication fee’
The peer review at PLoS ONE is the right question: is this publishable? Is the science sound? Not who might publish it or how can it be better rather only the simpler and direct questions. There is no other filter for quality. This moves good science forward as quickly as possible.
Was it a success then? After four years, they are the largest journal in the world in 2010. They’ve great community acceptance with 50,000 authors and 1,000 academic editors.
PLoS is promoting a paradigm shift.
Putting research in context: evaluating the article after publication, for its impact, by measuring: citations, web usage, expert ratings, social bookmarking, community rating, media and blog coverage, commenting activity.
Article-level metrics are not only about citations and usage, but a whole range of measures which are not only useful for quality evaluation but also for filtering and discovering content. PLoS hope other publishers follow suit.
Have have these data been received by authors and readers? very well, Peter shares several reviews. At plos-alm.opensci.info and in other visualizations, he also shares various views of data. Remember the slides of this presentation are at: http://tiny.cc/ALM12
With enough commenting data, even that can be evaluated semantically or otherwise to classify and better understand commentary. Aggregating per author is even more useful to some than data per article, as with the Association for Computer Machinery. http://tiny.cc/ALM8
What metrics are missing? Expert ratings, predictive metrics such as identifying guaranteed winners, media coverage such as in the Guardian or at MSNBC.com, better usage metrics and tracking conversations outside the publisher such as in Twitter or FriendFeed. Summary reputation metrics per commenter would also be useful.
To be done — filtering, navigation tools, an API at PLoS, more data sources and perhaps de-duplication between overlapping data sets in citations. They also want more expert analysis which they hope will be crowd sourced, and they want standards for measuring metrics between publishers. They’ve considered indexing, having one number to represent all others. They need for tenure committees to value the metrics, for them to be in resumes and to be considered by other publishers.
Article level metrics could be the start of an important new development in academic publishing.
Stephen Friend, Sage Bionetworks, Fred Hutchinson Cancer Research Center speaks on an effort just a few months old:
Distributed Tasks and Evolving Disease Models – to create an open access, integrative bionetwork evolved by contributor scientists and patients working to eliminate human disease.
Major, recent changes:
Sage believes disease biology should be open, where data are shared.
for more from Stephen Friend, see
http://sagebase.org/NEWSINFO/NewsInfoDownloads/Friend_BioIT_OCT09.pdf
Heather Joseph, Executive Director, SPARC, a three-soul team
SPARC means “Open Access” as access to journals and in digital form. SPARC is a library membership organization, founded to expand the dissemination of research results by using new technology. They seek especially to reduce financial pressures on institutions, and journals are a particularly fruitful area.
Why Open Access
The Budapest Open Access Initiative definition of Open Access: the primary concern is that knowledge should be freely available on the public Internet, but also usability is a concern — ability to read, download, copy, distribute, print, and so on.
How?
We often hear that “science is for the scientists” as an argument against Open Access: not only is that patronizing, it is untrue when you see the numbers showing demand for access to full scientific articles.
Another SPARC initiative was in copyright education: authors’ rights and the Scholar’s Copyright Project, SPARC + Science Commons
Harvard’s recent move toward institutional Open Access was historic — freely available articles regardless of publishing in journals, as a matter of policy. MIT and Stanford followed soon after, in the U.S. Faculties especially are recognizing the practicality of Open Access and leading the way.
see: The ROARMAP (Registry of Open Access Repository Material Archiving Policies)
Open Access as a policy concern: Alliance for Taxpayer Access — “American taxpayers are entitled to the research they’ve paid for.” — comprised of libraries, student groups, universities, and many others. They’ve stated and coalesced around four principles. The NIH public access policy — 12 months’ embargo, but far better than nothing and a proof of concept.
Currently, there is the Federal Research Public Access Act FRPAA, being reintroduced and the OSTP Public Access Policy Forum is providing the perfect environment for success. The Obama Administration is particularly interested in Open Access to federally funded research.
The opposition to Open Access is evolving:
The progress of SPARC and Open Access, in summary: infrastructure, options, acceptance, education, policies, and finally, the younger generation is rooting for Open Access.
Peter Murray-Rust is up, from The Open Knowledge Foundation and Unilever Centre at Cambridge
PMR returns several times to “Why Open Data is essential.” What data are open? Definitively, PMR says if it has a button from the OKF which says it’s open, then it is — they’ve shared the criteria online. Open Access and Open Data are not the same thing! see those definitions.
PMR is passionate about software as an agent of political change — Chem4Word is noted, and is a Chemistry Add-in for Word 2007 (written in C# so it’s perhaps not possible in OpenOffice). The bottom-up Web and Open Knowledge might best be exemplified in BlueObelisk.org, a community creating software, data, and other resources with few if any organizational constraints or bureaucracy.
Text and data mining are the next great stretch in the information age, when we understand not individual words but concepts communicated in normal human language. In linked Open Data there must be “zero friction.” Consider CrystalEye and the near-zero cost of robots in some cases.
The biggest source of friction, currently is legal, intellectual property. PMR points again to the Panton Principles as one remedy.
eTheses are the biggest, best source missing now in Open Science. How can librarians get these out there and open and faster? even immediately?
All the software PMR uses is open source, provided by the community Blue Obelisk. You can take Open Source software and innovate, which you cannot do with commercial software.
Prof. Richard Whitby’s Dial-A-Molecule (he’s at Southampton) — has a 20-year vision in which machines can visualize, can design and build molecules to spec with 100% success.
Perhaps 2 or 3 million reactions are published annually. Many published reactions and compounds are repeated — from journals, theses, and so on. Most journals are not free, are heavily copyrighted. Large organizations want to own data and that’s not going away. As such we need to develop agents of cultural change, some carrot and some stick, to get away from centralized gatekeeping.
WhatDoTheyKnow.org — the British Library was charging for access to Open Access publications, and via that site they are, in the UK required to respond to questions due to freedom of information laws.
Murray-Rust has started, since messaging journals such as J. Cheminformatics to ask about the Openness of their associated data. He started with Open Access journals and will work toward the more difficult journals.
They are interested in putting an Open Data button on their site. That in itself is a software based agent of social change, as an obligation.
CrystalEye — built by one grad student in under a year, it trolls every online publisher of crystal structures who do not hide their structures behind firewalls.
You can see Open Data and the software is Open Source, connecting then to definitions of Open Knowledge, and Open Services. See http://www.opendefinition.org/okd/ and http://www.opendefinition.org/ossd/ which are in use here, at the Open Science Foundation.
Antony Williams presents ChemSpider.com — free to use and open to everyone. It is a great step toward Open Access and Open Data. They’ve also an open source applet with which you can drill into data if the data are open.
Where is Chemistry online today? Still in need of a structure-centric community, though it was worse. Searches generally limited to text-based searches, dirty data, uncertain trustworthiness of sources , and too many searches required to find data.
People want fewer interfaces, more integrated. The semantic web for chemistry is already in place, and crowdsourcing generally is underway. Chemists will soon be able to search more by structure and substructure — in their own language. Data are more integrated, and the world is moving quickly toward Open Access and Open Data.
Classical business models will have to change.
Williams discusses cleanup of ChemSpider data via crowd curation, and integration of open knowledge such as portions of current Wikipedia articles. They’ve nearly released structure based results mixed with other properties, and with links through to papers, chemical suppliers, other data, related reactions, and so on.
Chemistry on the net is messy — there are pranks and then there are simply incorrect data, sometimes dangerous, or data which should be linked and are not.
PubChem for example — is an excellent platform, but they do not moderate the data and it’s messy. How would you clean such a large dataset? C&E News from the ACS has at times, Williams demonstrates, taken incorrect data directly from Wikipedia.
ChemSpider inherits dirty data, then takes the task of cleaning it. Expert and respected sources are wrong too often.
Oscar: a project which is the basis for semantic markup — entity-extraction, markup, and annotation.
ChemMantis from ChemSpider — finding all the chemical names and linking them out to ChemSpider, which connects to the whole web of data.
Embedding means never having to draw a molecule incorrectly, because there is a ready and embeddable reference at ChemSpider — “ChemSpider Everywhere” they call it. Spectra EMBED works in this vein, and they’ve even mobile apps for ChemSpider.
ChemSpider is not Open Source, but they use Open Source components. They run on SQL Server with licenses from Microsoft, for speed Williams says. They use OpenBabel, JSpecView, Jmol. The data itself is not Open Data though it is free to use. ChemSpider is a community resource but cannot make everything open without harming business models of algorithm providers.
Left to do? Plenty. Millions of depositions remain, a vast and ongoing curation effort, integrating RSC content which is a massive archive, and more.
ChemSpider is owned by the RSC, now and it began in a basement on a terribly low budget and a lot of effort. There are still only three full time employees. They are determined to keep use of ChemSpider free, and the RSC is a civil society organizations so they are focused on just sustainability, financially.
See on Twitter: @ChemSpiderman
Jean-Claude Bradley presents next on Using Free Hosted Web 2.0 Tools for Open Notebook Science (ONS)
Why? and other questions about ONS:
Bradley’s slides are online at: http://www.slideshare.net/jcbradley/nitle-open-notebook-science-talk
I’ll publish this post now for anyone remote and reading who wants to see those slides.
Bradley is demonstrating the wisdom of the crowd, or in his words “the blogosphere came to the rescue” at times when he did not know an answer or explanation and had shared work openly, making collaboration possible.
Read “The Telephone Gambit“ on the scandal of Bell’s lab notebook, it’s a book now freely available online, for the importance of open notebooks.
ONS is faster science, better science — go to Wikipedia to start reading up:
Being clear about how open you are: choosing wisely from these badges per their criteria http://onsclaims.wikispaces.com/ and stating publicly through use of these on your web site how open is your notebook.
Moving away from trust and toward proof — trust is not the issue when evidence is the test.
Bradley likes wikispaces.com for an open log, a nice and freely hosted service. Logging openly makes assumptions explicit, helps you see what you do not know. It also makes the rationale of your finding more explicit and thereby reduces assumptions, improving communications.
Use a video to share instructions which would take much longer to put into words, and put it on YouTube to do so openly. It’s free and maximizes its potential use and benefit.
Google Spreadsheets keep versions like a wiki, so you can restore an earlier version. Wikispaces provides good comparison of text versions, making it clear who has written or responded to given text.
Side benefits of being open: data uploaded to ChemSpider have later been turned into a game, spectralgame.com which enables further and crowdsourced science, impossible had data not first been open.
Sponsors of an ONS challenge have included: Submeta, Aldrich Chemistry, Nature — http://onschallenge.wikispaces.com/judges
Through approaches to improved search, more of science can be automated with bots and code.
Current issues in ONS include archiving open notebooks. With Andrew Lang, they’ve written some code and with Windows Scheduler to address daily backups of open notebooks. Google Spreadsheets allow downloading as MS Excel, which captures enough of the information to be a useful archive. They have periodic snapshots of everything, and share it at cost through Lulu.com via data disk (about $5). Wikispaces can export an entire space as HTML, they backup data files manually and then an automated process finishes the job.
“You are free” Cameron Neylon begins, “You are free to take notes, free to think, free to disagree with me.”
In turn, he next shows a slide with a good 100 or more names on it, crediting others for many of the ideas he has or will share. The point, clearly is that he is free and open in his thought and sharing and we should be, too.
His slides, similar at least to what he presents today are at http://www.slideshare.net/cameronneylon and specifically http://www.slideshare.net/CameronNeylon/science-in-the-open-science-commons-pacific-northwest
Cameron reviews the reasons society funds science and the reasons he pursues it himself, in both cases noting those from practical to whimsical. He notes that his conduct of science is a privilege, not a right. How does he deliver on his obligation to the public who fund him? Though not in pure economic terms, science must maximize the return on the public’s investment.
Easy to say, but how? Open Access — making sure results are available, to leverage the largest possible community for building further on results. Cameron flatly states that the traditional process of submitting, reviewing, rejecting papers in peer reviewed publishing is overkill, not always but often.
Backing up a bit, he demonstrates the paucity of decent search results for basic information such as the solubility of boc-Glycine in ethanol. Then, he shows a quality search result for same thanks to Jean-Claude Bradley and his conduct of Open Notebook Science.
Broadcasting on the web is easy. Sharing it effectively is harder. You’ve got to share work in a form others can use, whether peers or students or the interested public. Technical interoperability and legal interoperability can be equally vexing. On the later, Creative Commons and Science Commons have come to understand that legal terms must be as specific as possible, and they define and offer licensing accordingly.
Panton Principles — Principles for Open Data in Science — http://pantonprinciples.org
Differences between the Open Knowledge Foundation and Science Commons were set aside in a pub, and they found common ground: data related to published science should be explicitly placed in the public domain.
Mapping a process onto agreed vocabularies and building software to do those things, only and at a time, rather than “control panel” type software designed logically for every contingency but with no consideration for human limitations.
Science needs tools which capture inherent data structure. Cameron works in Google Wave, perhaps the most open with respect to time, in that it is the most real-time of any collaboration tool available. He demonstrates Wave’s ability, via plugin or bot to generate RDF automatically from his streaming, open notebook. (I snapped a photo of Google Wave being demonstrated at Microsoft Research; will share it later if it turns out.)
Is it possible to share the entire research record? Yes, the technology exists. Do scientists want to share? Most do not.
The public and their governments want more open and better science. How to persuade scientists?
See Cameron’s show of “Submissions to Genbank” — the scale problem is so bad, for data here and in most fields of science, that tools like Genbank are failing. People of course, do not scale — i.e. are not getting much smarter. Governments do not scale, either, being human. Research groups, also do not effectively scale, also being human.
The web does scale. Scientists will need to be web-native just to survive, to keep up with Big Data.
The network requires open content for its foundation; science must connect, must be open.
The Fourth Paradigm — Data-Intensive Scientific Discovery – the first Creative Commons licensed publication from Microsoft Research
How is computer science transforming science? Through four pillars: health and well-being, scholarly communications, earth and the environment, and core computer science.
They wanted from the get-go to honor Jim Gray, which included that the work should be open and freely available to everyone. In early 2007 he introduced the “Fourth Paradigm of Scientific Discovery.”
Microsoft’s External Research group has had several partnerships with Creative Commons, including inserting Creative Commons licenses in Office 2007 and an ontology add-in (importing and embedding an ontology) for Word 2007 with XML mark-up.
We did have the FriendFeed room included here, too but when FF went down it was dragging this blog with it, so I removed it.
Looks like the live stream and chat are up and running at http://chris.pirillo.com/live/ and with 400+ viewers at the moment. Great stuff!
http://chris.pirillo.com/live/ and with 40',description:'Looks like the live stream and chat are up and running at http://chris.pirillo.com/live/ and with 40'})"> Thanks Mike, for sharing the image I used behind this blog’s header. It seemed just about right; optical fiber owes its success to science and engineering alike and is most known for use in communications. Light, networking, and even the bokeh ends that look like venn diagrams, all fitting for the symposium.
Most fitting is how generous Mike is with his work, heck he’s a Creative Commons case study. Below, enjoy Mike’s video version of the fiber. BG
Flickr Video
p.s. This was created as part of my work with IPAS – the Institute of Photonics and Advanced Sensing led by Prof. Tanya Monro.
Way cool Mike and thanks for chiming in.
Bloody excellent summary, Brian !! Thanks a bunch for taking the time to do all of this. I stayed tuned to the livestream till about 01:30 am !!
@Mike, we meet again 
@Brian, here’s a neat related podcast from 2008 about open science hosted by Mike & Dave Wallace:- http://extraordinary.thepodcastnetwork.com/2008/07/28/extraordinary-everyday-lives-053-open-science/
Glad you found a use for it. Hat tip to http://twitter.com/McDawg for a tweet pointing me to this post. Listening to the ustream recording of Cameron Neylon right now as he says ‘standing on the tweets, blog posts, … shoulders of giants’ – That’s the point of granting permission in advance.
Gave a great #scspn day.
@fang – Mike Seyfang