Please follow this easily memorable link:
“Biased ART: A Neural Architecture that Shifts Attention Toward Previously Disregarded Features Following an Incorrect Prediction,” by Gail A. Carpenter and Sai Chaitanya Gaddam was the most viewed article on BU Digital Common last month. The technical report was added to DIgital Common on November 11 and was viewed 42 times during the rest of the month.
Abstract: Memories in Adaptive Resonance Theory (ART) networks are based on matched patterns that focus attention on those portions of bottom-up inputs that match active top-down expectations. While this learning strategy has proved successful for both brain models and applications, computational examples show that attention to early critical features may later distort memory representations during online fast learning. For supervised learning, biased ARTMAP (bARTMAP) solves the problem of over-emphasis on early critical features by directing attention away from previously attended features after the system makes a predictive error. Small-scale, hand-computed analog and binary examples illustrate key model dynamics. Twodimensional simulation examples demonstrate the evolution of bARTMAP memories as they are learned online. Benchmark simulations show that featural biasing also improves performance on large-scale examples. One example, which predicts movie genres and is based, in part, on the Netflix Prize database, was developed for this project. Both first principles and consistent performance improvements on all simulation studies suggest that featural biasing should be incorporated by default in all ARTMAP systems. Benchmark datasets and bARTMAP code are available from the CNS Technology Lab Website: http://techlab.bu.edu/bART/.
Digital Common, the University’s open access repository hosts a wide range of materials including published articles from faculty, technical reports, theses and dissertations, learning objects, and items digitized by the Libraries. For additional information about the repository, contact Vika Zafrin (firstname.lastname@example.org), the Institutional Repository Librarian for the BU Libraries.
This year, BU is again participating in Open Access Week, an international event sponsored by the Scholarly Publishing & Academic Resources Coalition (SPARC). BU Libraries will sponsor a Q&A with Peter Suber, widely regarded as one of the originators of the Open Access movement.
WHO WHAT WHERE WHEN
The conversation will take place WEDNESDAY, OCTOBER 26 at 4PM in Stone B50 (basement of the CAS building, 675/685 Commonwealth Ave.)
Peter Suber is a Fellow at Harvard’s Berkman Center for Internet and Society, Senior Researcher at SPARC, the Open Access Project Director at the Public Knowledge Project, and Research Professor of Philosophy at Earlham College. He earned his MA and PhD in philosophy, as well as his JD cum laude, from Northwestern University. Suber serves on numerous steering committees and advisory boards for high-profile open access projects. His homepage is here.
WHAT’S AT STAKE?
Open access is an important and growing area of concern among BU faculty, according to the 2010 Faculty Library Survey Report. The academic publishing industry is in the midst of a complex upheaval, as authors begin to pressure publishers to change copyright-related practices and institutional libraries buckle under astronomically increasing subscription prices for academic periodical literature. Many publishers have changed their standard practices to allow your work to be freely disseminated after a certain period following publication. Some others are in the process of implementing similar changes. Still others are holding fast to practices that ultimately obstruct the dissemination of knowledge.
HOW ARE BU FACULTY AFFECTED?
If you have published articles, and intend to do so again, you are affected. At stake is whether your work will effectively reach your audience. As an author, you have a say in this. BU has resources to help clarify your rights, the current state of academic publishing, and venues for disseminating your research.
You are affected as a reader, as well. Open access directly influences how much material is available to you for research, irrespective of library budget constraints.
GET YOUR QUESTIONS ANSWERED
You spoke, and BU listened. The 2010 Faculty Library Survey Report indicates that there is a growing interest in open access among our faculty. Please come with questions about open access and what it means for you and your research. After a brief introduction, we will open up the floor for the Q&A.
Sponsored by BU Libraries’ Digital Initiatives and Open Access Group
Rick Kulkarni, MD, Medical Director, eMedicine.com, and Assistant Professor of Surgery, Yale University School of Medicine, discusses open-access publishing.
A group of herpetologists—researchers who study reptiles and amphibians—has been quietly demonstrating that it’s possible to put together a well-regarded, researcher-run journal with the tiniest of budgets and no help from a publisher.
The journal, Herpetological Conservation and Biology, caught my eye as a well-developed example of a movement for grass-roots scholarly publishing that has been rapidly picking up speed. The herpetology publication, founded in 2006, is an online-only, open-access, peer-reviewed journal with a budget of about $100 a year. (That money comes out of the editors’ pockets.) Unlike most science journals, it charges no author or download fees. It has a submission-to-publication turnaround time measured in weeks or at most a few months.
And it has just hit a milestone: The editors learned in December 2010 that HCB will be included in Journal Citation Reports, a service run by the commercial publisher Thomson Reuters that calculates impact factors for journals—a significant measure of importance for many researchers. HCB will receive its first impact rating in 2012 or 2013, and the editors expect the journal to rate highly. That credential will help reassure potential contributors, especially researchers who don’t yet have tenure, that publishing an article in HCB will be good for their careers.
© 2010 Brad Wheeler. The text of this article is licensed under the Creative Commons Attribution 3.0 Unported License (http://creativecommons.org/licenses/by/3.0/).
EDUCAUSE Review, vol. 45, no. 6 (November/December 2010)
Brad Wheeler (email@example.com) is the Vice President for Information Technology and CIO for Indiana University
The economic crisis that began in 2008 is prompting a large-scale search for cost savings in colleges and universities. For example, a number of institutions have engaged management consultant firms (e.g., Bain & Company, Accenture, Hackett Group) to help them rationalize business processes, benchmark costs, and seek economic efficiencies of the type that are presumed in the private sector. In most cases, public reports from this work point to potential efficiency gains for basic business practices such as centralizing procurement, making greater use of common IT systems, and reducing layers of management. The ultimate goals of these efficiency exercises are to redirect more of an institution’s resources into its core missions of research and education and to spend less on basic, non-core, administrative-support activities. However, since the largest proportion of the budgets of higher education institutions is (appropriately) in the core areas, gains in both efficiency and effectiveness are essential there as well if institutions are to adjust to the economic realities of this decade.1
Renting the Copier
Imagine if an efficiency consultant uncovered the following situation. In 1984, an important business function in a university began using a convenient copy service for $0.10 per copy. The university staff provided most of the labor, and the copy service company provided the machine. Twenty-six years later, the university is still using the copy service and still providing most of the labor costs, but the price has changed—to $0.85 per copy. In addition, the copy service now imposes restrictions such that only certain members of the university can read the copied documents. If others want to read the copies, the university must pay extra for the number of potential readers of the document rather than just the per-copy fee.
Ridiculous? A fee that is 850 percent of the rate twenty-five years earlier? The fee should be approximately $0.21 per copy if it had tracked with the U.S. inflation rate and had gained no technology efficiencies in twenty-six years. Surely, no efficiency expert would affirm continuing to rent the copier under these terms. The expert would undoubtedly point out that the university could own and operate a copier for a much lower cost and without any restrictive use rules that impede the university’s work. If this situation were found in a non-core administrative area, it would be viewed as outrageous and would be changed immediately.
Renting the “Journal Copier”
In fact, a very similar situation exists today in a core area of research and education. This copier example is imperfectly illustrative of the cost and restrictive use imposed on most academic libraries by academic journals. For example, in 2006 a group of 275 doctoral/research universities paid a combined $1 billion to essentially rent the “journal copier” system that would provide their faculty, staff, and students with access to scholarly journals.2 In 2010, they are paying even more, and the real scale of the cost to colleges and universities spans globally to institutions of all sizes. A sweeping study from the United Kingdom estimated total annual expenditure for journals at £597.4 million (approximately U.S. $952 million) in 2006–7. The total estimate for scholarly communications—inclusive of faculty time for editing and reviewing—was £5.4 billion (approximately U.S. $8.6 billion).3
Librarians and others have long chronicled the “Crisis in Scholarly Communication,” so I will not repeat it here.4 The move from libraries owning paper copies of journals to paying for electronic access via annual subscriptions—that is, “renting” the material—has shifted absolute pricing power to the publishers. Prices have steadily escalated even as electronic distribution has cut production costs and as colleges and universities continue to supply most of the labor and intellectual skills required for journal writing, reviewing, and editing.
A number of solutions have been proposed for this cost and access problem: the assertion by faculty authors of their rights to deposit work in their institution’s library; models of subsidized open-access publication; and federal open-access mandates for sponsored research. All of these ideas have great merit, and in time, they could favorably affect the $1B problem. I fully endorse them as worthy directions. In the near term, however, I see little real relief for the inflating $1B problem without a more substantial intervention. The economics of the relationship between the buyers (fragmented) and the suppliers (consolidated) are vastly skewed and show no plausible reason for self-correction.
Since scholarly communications span many faculty, institutions, and continents, no college or university can influence the $1B cost problem by acting alone. Few journals are owned by institutions. Many are owned by scholarly associations (e.g., the American Psychological Association), which source the digital production of their journals to a diminishing number of commercial publishing firms (e.g., Elsevier, Wiley). In the sciences, for example, many journals are owned entirely by the commercial publishers. The scholarly publishing industry is thus composed of many fragmented participants and agency relationships for authoring (faculty), buying (libraries), and consuming (faculty/students). There is no locus of decision-making authority among institutions to produce a directed solution to the $1B problem.
How can CIOs and other IT leaders help? A new model is needed—one that uses a low-cost, IT-enabled infrastructure to incent industry evolution to a cost structure and rules that better enable the work of scholars in a financially sustainable way.
The Big Digital Machine
If exorbitant prices for “renting” the journal production infrastructure and restrictive rules of access are the problem, then institutions should shift from renting to owning the “journal copier—or leasing it under more favorable terms. The Big Digital Machine (BDM) is a concept to do precisely that. It envisions aggregating and integrating a set of publishing capabilities for the full cycle of scholarly publishing—from production to distribution to preservation. Its services would include preprints, conference proceedings, journal articles and journals, monographs, and textbooks. It could make a better offer to the many scholarly societies that have outsourced their journal publishing work to commercial firms. By collectively owning BDM, institutions could set the terms for journal pricing and distribution, and they could better manage the costs of the infrastructure—indirectly paid by colleges and universities—on which scholarly societies run their review and publication processes. BDM could be the “back office” system for both existing journal titles and new journals. I assert five reasons for pursuing BDM:
- Libraries are nearing (if they are not already beyond) their capacity to cut and prune in all directions to pay for the rising costs of journals. Faculty have long been shielded from these problems, but now the costs of the current model mean that institutions are cutting access to the scholarly record in some fields. This undermines the research mission of the institutions, and in the “Era of the New Normal,” there is no new money to maintain the status quo as prices continue to rise.
- We stand as the first generation in a millennium that cannot guarantee access to scholarship for future generations. When copyright is transferred from author to publisher in the current model, institutions must then rent it back each year to read their own scholarship—at whatever price is asked. Ownership means that societies or institutions can set their own terms for access to their scholarship—from open access to subscriptions, per-article pricing, and everything in between.
- We have already proven much of the technology and systems that are needed to create BDM. Tools like Open Journal Systems (OJS), Fedora/DSpace/DuraSpace, Connexions, EPrints, libraries’ institutional repositories, and others are fulfilling many of the core capabilities that are needed for a college/university-owned publishing infrastructure.
- Colleges and universities can aggregate their resources and run above-campus or cloud service models that serve multiple institutions.
- Ownership of the infrastructure means that colleges and universities can make a better offer to the scholarly societies and other journal providers to achieve their goals using BDM. Once BDM is established, it will be difficult to further justify exorbitant rental fees for journals that choose other approaches.
A Modest Start
In 2009, the institutions in the Committee on Institutional Cooperation (http://www.cic.net) collectively invested $160,000 in early work on the BDM concept. That work developed some technical integration among DuraCloud, OJS, and Connexions software. In addition, these software projects agreed to adopt a common approach to Google Analytics so that authors could assess the use of their content in a uniform way. By the summer of 2010, much of this work was being rolled out in new releases of the various systems.
As of this writing, the next steps for the BDM concept are not determined, but the rising costs and the use restrictions of the current model remain unabated. I believe that visionary leaders can solve the $1B problem that affects the core of the academic mission, but doing so will require the scale of coordinated and visionary action that created Internet2 more than a decade ago.
So, after those management and efficiency consultants help colleges and universities streamline procurement and other non-core activities, we need to turn their attention to fixing one of the big money problems in the core of the academy. Or we could just get started on it ourselves.
1. See William F. Massy, “It’s Time to Improve Academic, Not Just Administrative, Productivity,” Chronicle of Higher Education, January 9, 2009, p. A26, <http://chronicle.com/article/Its-Time-to-Improve-Academ/2222/>.
2. Tai Phan, Laura Hardesty, Cindy Sheckells, and Denise Davis, Academic Libraries: 2008, NCES 2010–348 (Washington, D.C.: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education, 2009), p. 13, <http://nces.ed.gov/pubs2010/2010348.pdf>.
3. See John Houghton et al., “Economic Implications of Alternative Scholarly Publishing Models: Exploring the Costs and Benefits,” JISC report, January 27, 2009), pp. XIII, XI, <http://www.jisc.ac.uk/publications/reports/2009/economicpublishingmodelsfinalreport.aspx>.
4. See Michigan State University Libraries, “The Crisis in Scholarly Communication,” <http://www.lib.msu.edu/features/crisis/>, and University of Connecticut, University Libraries, “Scholarly Communication Crisis,” <http://www.lib.uconn.edu/about/publications/scholarlycommunication.html>.
Harvard University Press has recently published Marcus Boon’s book, In Praise of Copying, under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported license, and it is available as a hardcover as well as a freely available PDF file. Below is a brief excerpt:
Boon, Marcus. 2010. In praise of copying. Cambridge, Mass: Harvard University Press.
I know that people reading this book will expect to find here an ethics of copying—but from the outset, I would like to call such a desire into question. Can we really identify an area of human activity outside copying which would make it possible for us to choose or decide whether to copy or not? I will argue that there is no such area, that we are always entangled in the dynamics of mimesis, and I write “in praise of copying” as an affirmation of copying rather than as an ethics. The word “copyright” (nearly 3.8 billion hits on Google) itself sounds a little desperate, as though one had to actually suture the words “copy” and “right” together in order for them to associate consistently. Just to put that number in perspective, “freedom” gets only 315 million Google hits and “truth” 312 million—a factor of ten less than “copyright.” Even “sex” gets only 876 million hits, in case you’re wondering. Don’t you think that the concept of “copy- right” is a little overdetermined? (p. 6)
For Immediate Release
October 21, 2010
Boston University Libraries are pleased to announce that Kathleen Fitzpatrick is the speaker for the 2010 Fall Lecture on Open Access to be presented at 3:00 pm on October 29, 2010.
Professor, Department of Media Studies, Pomona College
|Topic:||Planned Obsolescence: Publishing, Technology, and the Future of the Academy|
|Date:||October 29, 2010|
|Place:||Photonics #206, 8 St. Mary Street, Boston, MA (map)|
Professor Fitzpatrick has received many fellowships, grants, and awards. In addition to numerous articles and media projects, Fitzpatrick is Co-coordinating Editor and Press Director, MediaCommons and author of The Anxiety of Obsolescence: The American Novel in the Age of Television. Nashville, TN: Vanderbilt University Press, 2006.
- Named an “Outstanding Academic Title” by Choice, the publication of the Association of College and Research Libraries, January 2008.
- Selected as a “book of the month” by the Resource Center for Cyberculture Studies.
Fitzpatrick is currently working on a book-length project focusing on the social and institutional changes necessary to developing the digital future of scholarly publishing, under contract to New York University Press. Manuscript completed; undergoing second-round review. Available for open peer review online here.
… Though the notion of a crisis in scholarly publishing was first aired well over a decade ago (one might see Sanford Thatcher’s 1995 article in the Chronicle of Higher Education, entitled “The Crisis in Scholarly Communication”), things suddenly got much, much worse after the first dot-com bubble burst in 2000. During this dramatic turn in the stock market, when numerous university endowments went into free fall (a moment that, in retrospect, seems like mere foreshadowing), two academic units whose budgets took among the hardest hits were university presses and university libraries. And the cuts in funding for libraries represented a further budget cut for presses, as numerous libraries, already straining under the exponentially rising costs of journals, especially in the sciences, managed the cutbacks by reducing the number of monographs they purchased.
Planned Obsolescence, “Introduction”
Originally posted to Scholarly Communications @ Duke
by Kevin Smith
October 21, 2010
Yesterday I was part of a fascinating discussion between librarians and a faculty member who has just become head of the publications board of her scholarly scientific society. She was anxious to learn from librarians about how we approaching bundling deals, pricing policies and open access. Lots of this conversation was really interesting and important for both parties, but the biggest thing I took away from the discussion was the faculty editor’s desire to convince her large commercial publisher to make open access options more available and less costly for her authors. She understood that open access would benefit those authors and increase the journals’ impact, thus helping to ensure their long-term viability. She also saw through the publisher’s hype and grasped that they were really trying to inhibit open access with their policies, not facilitate it.
It was in the context of that discussion that I finally got around to reading the new Journal Publishing Agreement from the American Chemical Society. The hype from ACS is that their new contract “expands author rights and clarifies author responsibilities.” The reality is that ACS authors will continue to be behind the curve in scholarly communications under this agreement.
I have not compared the old and new agreement closely, so I am prepared to accept that this agreement is an improvement over the old one for authors. It does outline quite clearly, in some cases, what authors can do with their own work once they cede copyright, gratis, to the ACS.
In this post on the Book of Trogool blog, Beth Brown does a nice job of explaining some of the problems with the new ACS agreement. I will amplify some of her points here and also add one very picky and lawyerly note.
Most importantly, of course, the new agreement is a complete transfer of copyright to the publisher. Sometimes authors think they might still hold copyright in earlier versions of their articles, written prior to what they submit. This is probably incorrect as a legal matter, but the ACS is taking no chances. The new agreement specifies that ACS will hold the copyright in all versions of a submitted article. As Brown indicates in her post, and our conversation with our faculty editor also proved, this is moving against the trend in most other sciences and shows why chemists will continue to be disadvantaged by their scholarly society.
Of course, if authors retain sufficient rights to disseminate their work, this problem could be alleviated. But in all three important ares, this is not the case.
First, for authors using their own work with students, that new JPA encourages use of a link to their “articles on request” service. If authors use that, their students will have access to 50 free downloads of the article and, after that, will have to pay a per-use charge. After one year the article on request link will provide free access. So there is a pretty clear attempt to generate some new income by getting students to pay for articles outside of the publisher’s subscription income stream. In fairness, however, this is the first listed method of using work with students but not the only one; authors are allowed to post their articles in secure systems for student access.
Next, distribution to colleagues is also limited, this time entirely, to the “articles on request” service described above. Thus authors essentially get 50 free “offprints.” After that someone, either the author herself or the colleague with whom she wants to share her work, must pay. This pay wall is supposed to disappear after 12 months, but by then the window for productive sharing has probably closed.
Lastly, authors’ ability to put their work into an open access repository is severely limited. It looks at first glance like authors have pretty broad permission to post their submitted version of their article, but there is a condition. Authors must get written permission from the journal editor that posting of their submitted version does not violate the policies of the particular journal. So this attempt to be clearer about author rights really hides behind the obscurity of potentially various journal demands. When we get to posting the published version, this is allowed only if the author is subject to a mandate, either from a funder or the author’s institution. These mandates can be complied with after a 12 month delay. If the mandate requires quicker access, the author is told they must pay for open access through the ACS “Author’s Choice” program.
It is interesting that this new policy actually creates an incentive for open access mandates. I’m not sure its authors were really thinking that one through; they just saw OA as a threat and moved to postpone it for 12 months, to the ultimate detriment, we can be sure, of chemistry authors and researchers. Another place the JPA authors were not paying close attention — and this is the nitpicking I promised — was in the clause on “supporting information” that accompanies an article, for which, the agreement says, the copyright transfer is non-exclusive. This is simply poor drafting, since a non-exclusive transfer is essentially an oxymoron. The agreement goes on to clarify that both the ACS and the author have full rights to exercise copyright in such material, which is more or less a “joint authorship” situation. I puzzled over this odd clause with another lawyer yesterday, and our conclusion was that the actual effect of this language, if a court had to interpret it, would probably be a transfer of copyright to the ACS and a license back to the author. But whether that licenses is itself exclusive or non-exclusive and the exact scope of it is unclear.
This new ACS agreement clearly indicates a desire to stave off open access and to control it in a way that does not threaten the traditional thinking in scholarly publishing. But as our faculty editor indicated yesterday, this traditional thinking is no longer good enough. In her field, she told us, open access journals are arising that pose serious competition to the more traditional journals that she is responsible for. She is already afraid that these OA journals will grow in success and that her journals, with their very limited OA options, will suffer. The same fear ought to grip the ACS. Sooner or later chemists will find new ways to disseminate their research, just as other scientists have already done, and ACS journals will, without further change, begin to decline. OA is a proven benefit to scientific research, and this attempt by ACS to grasp and control it so tightly might just backfire.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 United States.
In keeping with this month’s Open Access theme, two events are taking place at BU. All are invited to these important discussions.
Missed Connections: Research Barriers, an Interactive Discussion
with Dan Benedetti, BU Libraries
Wednesday, October 20, 3pm
Howard Thurman Center (Lower level GSU)
Planned Obsolescence: Publishing, Technology, and the Future of the Academy
Kathleen Fitzpatrick, Media Studies, Pomona College
Friday, October 29, 3-5pm