INRIA’s Literature Goes Open Access
Document Actions
|
Last October, ten major French scientific organizations (1), among which INRIA, signed an agreement whereby, each will contribute to a common on-line and open access scientific publication platform. The French universities are also expected to climb on board soon. Rather than starting from scratch, it will build upon HAL (Hyper Article en Ligne), a web database initiated by CNRS-run CCSD (2), as early as 2000. This server already comprises some 40,000 documents, with INRIA sub-portal accounting for 11,000 out of this total. Illustrating its institutional commitment to the open archives, INRIA has contributed its entire collection (3). In addition, researchers are invited to deposit their scientific output. Self-archiving on HAL is already a habit for roughly one third of them. The institute’s goal is to reach 100% within a 5-year period. In the academe of yore, scholars used to rely mostly on private companies for their scientific publications. While discharging scientists from the toilsome task of having their books printed and retailed worldwide, publishers would also establish prestigious and nearly inescapable collections of academic journals, in fields ranging from medicine to nuclear physics, or applied mechanics. The advent of the Internet and its promise of discloads of freely accessible information have prompted the research community to reconsider this situation. Freely accessible? Not quite. E-prints from commercial publishers come to a price. Many libraries and institutions complain subscription fees have skyrocketed to such stratospheric heights (4), they can no longer purchase access to all the information that their researchers require. In the mean time, publishers entered a merger frenzy resulting in a marketplace dominated by ever fewer companies (5) with whom negotiation margins are fast shrinking (6). A handful of large publishing houses have tremendous market power, charging ever higher prices.
Subscribers’ first reaction was to try regain bargaining clout by grouping
into consortia whose purpose is to bulk buy for their members at the best price.
The French consortium Couperin, for instance, boasts over 200 members,
including 23 research institutions. Other voices in the research community send
broadsides against the current scholarly communication model. They argue that the research funded by public money is given away for
free to publishers who, eventually, will charge huge sums for letting scholars
access content they created in the first place. |
|
The open access groundswell came as a response to this longing for
reappropriation, with pro bono goal of providing freely accessible repositories
of intellectual material. It accounts for a paltry 20% the world output, but in
some domains like theorical physics, figures nudge 90%. “The
gist is for researchers to retake control of their own scientific production,
sums-up Pascale Laurent, head of Irisa documentation center. Until recently, this approach was very much
a matter of individual choice: a personal initiative-based movement epitomized
by a few scientists (7). It is only at a second stage that
scientific institutions also started getting on board, Laurent supplies. In 2003, in Berlin,
Germany, an appeal
was launched calling for these institutions to go open access. Major French
research institutes like CNRS, INRIA and Inserm where among first signatories.”
They have joined the CNRS-initiated HAL free access platform. With 1,200
texts added monthly, this repository of transdisciplinary material receives an average
15% of the national scientific output.
Operated on volunteer-only basis, HAL doesn’t require researchers to deposit their work in any mandatory fashion ...yet. But that may change in the future as DRGI, a department of French ministry of Research (8), considers “the option” of making compulsory the publication for results of research that have been financed through ANR, the French national research agency.
The same idea is being tossed around in Brussels. In a recent report (9), Eurab, the European Research Advisory Board (10) recommended that “the Commission should consider mandating all researchers funded under FP7 (11) to lodge their publications resulting from EC-funded research in an open access repository as soon as possible after publication, to be made openly accessible within 6 months at the latest. Authors should deposit post-prints (or publisher’s version if permitted) plus metadata of articles accepted for publication in peer-reviewed journals and international conference proceedings.” Furthermore, as a matter of policy, “the Commission should strongly encourage all Member States to promote open access publication policies for all their publicly funded research.”
The report follows a January 2006 survey (12) in favour of “a European policy mandating published articles arising from EC-funded research to be available after a given time period in open access archives. (…) This archiving could become a condition for funding.” Fearing a barrage by editors lobby, open archives advocates have started a petition (13) urging the Commission “to endorse the recommendations in full.” In particular, “as a matter of urgency” make open archiving a condition for funding. 20,000 scientists have signed so far. Further along in the pipe is the long-run goal of meshing together the scattered depositories of all member states, into a broader and fully standardized European knowledge base. Roadmap to this ambition is Driver: Digital Repository Infrastructure Vision for European (14). 10 universities, so far, have joined this partnership in which EU sees a content-oriented complementary counterpart to Géant2, the much appraised infrastructure for computing resources and data transport. |
|
From a
technical viewpoint, most open archives depositories (15)
comply with PMH (16), the emerging
technical standard in the digital library world. Among other things, this Protocol
for Metadata Harvesting enables authors to easily deposit their work on various
compatible platforms. “That is called
cross posting, Laurent explains. For instance, one can deposit on HAL and have it automatically available on another
sites like ArXiv.” Better
yet, all this XMLed raw material is fully crawlable, browsable and searchable
by a host of engines including popular tools like Google or more specialized
breeds of knowbots (17). Having said that, all depositories are not fully compatible yet. “We are still in an in-between situation, Laurent says. It will take an additional 10 years before things fully settle down.” Will publishers throw spanners in the works? “A conflict is nobody’s interest. We sure must avoid sending researchers on the warpath against publishers. That won’t bring any good. Both groups have their respective function and distinct vocation.” Furthermore, there is a motley assortment of policies out there. “Some publishers allow authors to post on open archives depositories, provided these are institutional depositories. Conversely, sometimes, researchers might not be permitted to post their work on their personal web site. The Sherpa platform has classed publishers into 4 categories (18) depending on the restrictions they imposed upon authors.” |
The number of postings on HAL is fast rising. |
On the other hand, even researchers do not all share blithe enthusiasm for OA. “When they decide to publish their work, authors do not pick the place randomly, Laurent explains. They select the review according to three criteria: its international fame, its impact factor (19) and its editorial committee (20). If there isn’t any peer reviewing, they will look askance at submitting. Quite often, before posting, they will be waiting until the paper has been published somewhere else first. ” Nonetheless, “one of HAL’s advantages resides in its capability of aggregating papers into an electronic so-called “overlay “reviews that are then peer-reviewed. But a precondition is to credibilize e-reviews. And no one else than the institutions can achieve such goal.” Also annoying to many, is the fact that, once a pre-print paper is posted on HAL, the author can’t delete it. “That’s a feature researchers dislike”, Laurent says. Other authors worry about copyright infringement on fresh material.
Lastly, some do not feel tremendously comfortable with the indicator aspect of the matter. In the above-mentioned report, EURAB opines that FP7 should contain “an action to invite proposals for an enhanced ranking of journals which includes not only traditional indicators of impact but also open access policies.” As a yardstick of research output quality, journal article publication counts and citations are of essence. Due to its sorting capabilities, HAL could well become a tool for fathoming output and measuring individual performances. “Measuring or monitoring? Laurent asks wittily. Some fear, HAL will help keep tabs on them. And they do not feel comfortable with the idea. French LOLF law (21) insists on having relevant indicators to ascertain the value of public money invested in research projects.” Things get even a bit trickier for there is an institutional dimension to the problem. “Authors and institutions have different logics. These latter are interested in the troop-counting. But this reckoning must be performed along a multitude of configurations and labels: region academic groupings, universities, faculties, laboratories, establishments, mixed units of research, university research units etc.” A bit of a headache as “an author can appear on an establishment payroll, while working for a lab and being part of a research team at the same time. Take researchers here: some mention the local Irisa, while others mention national Inria.” The grouping and branding of local universities under a new European University of Brittany (22) will add yet another layer of complexity to the matter.
Footnotes
(N1) The ten signatories are : Conference of
University Presidents (CPU), Conference of French Grandes Ecoles Directors (CDGE), Public Agricultural and
Environmental Research Institute (CEMAGREF), French Agricultural
Research Centre working for International Development (CIRAD), National Center for Scientific Research (CNRS),
French National Institute for Agricultural Research (INRA), French National institute for
Research in Computer Science and Control (INRIA), French Public Research
body dedicated to human health (INSERM), Institut PASTEUR, Research Institute for Development (IRD),
with the support of the Ministry for Research and Higher Education.
(N2) CNRS
(Centre national de la recherché scientifique) is the French National Center for Scientific Research. CCSD (Centre pour la communication scientifique directe) is
its Center for Direct Scientific Communication.
(N3) The Beijing-based joint-lab LIAMA tail-end-charlied a few
weeks ago, contributing an additional 80 documents. LIAMA is an operative
research hub created by INRIA and the Chinese Academy
of Sciences in 1997. Acronym
stands for: Laboratoire franco-chinois d'Informatique, d'Automatique et de
Mathématiques Appliquées.
(N4) According the Association of Research Libraries, prices for
reviews have increased by 215% between 1986 and 2001. An
illustration of price rise is this survey in the field of maths. Also of interest: this survey on the practice of bundling subscriptions to ejournals.
(N5) The three main scientific publishing groups in the world are: Reed Elsevier, Springer Science+Business Media and Taylor & Francis.
(N6) In 2006, the entire editorial board of the prestigious Topology journal resigned, in protest of the publisher’s pricing policy. “Since Elsevier gained control of the journal, (…) we believe that the price (…) has had a significant and damaging effect on Topology’s reputation in the mathematical research community.” Elsevier used to charge $1,665 for the yearly subscription. Secessionists plan to launch a new journal (Journal of Topology) for only $570, starting 2008.
(N7) Among open access advocates, two names have come
forward. Physicist Paul Ginsparg started his own brew of open archive server
back in the early 1990s. Now universally known as ArXiv, this repository currently hosts 400,000 e-prints in physics,
maths, computer science and so forth. The move is also epitomized by Hungarian
scientist Stevan Harnad who has been doing a lot of open-access archivangelism of late.
(N8) DRGI: Direction générale de la
recherche et de l'innovation. Directorate-General of for Research
and Innovation.
(N9) European Research Advisory Board, Final Report, Scientific Publication: Policy on Open Access. December 2006.
(N10) Eurab is: European Research Advisory Board. An advisory committee created by the Commission and made up of 45 experts, Eurab provides advice on EU research policy.
(N11) FP7. On 18 December 2006, the Council adopted decisions establishing the Seventh Framework Program of the EU for research and technological development for the period 2007 to 2013.
(N12) Study on the Economic and Technical Evolution of the Scientific
Publication Markets in Europe, Final Report,
January 2006.
(N13) The petition has collected over 24,000 signatures so far. In the wake of it, American scientists started their own petition, asking free and open access to research funded by the US federal government.
(N14) Driver:
Digital Repository Infrastructure Vision for European Research: A testbed for a repository infrastructure
that will enable researchers to plug into the new knowledge base and use
scientific content in a standardised, open way.
(N15) For a list of the world OA depositories check: Opendaor.org
(N16) OAI-PMH
stands for: Open Archives Initiative - Protocol for Metadata Harvesting. It is
the emerging technical standard in the digital library world. The Open Archives Initiative
promotes interoperability standards that aim to facilitate the efficient
dissemination of content. It has its roots in an effort to enhance access to e-print
archives as a means of increasing the availability of scholarly communication.
(N17) Among specialized bots is Openarchives.eu This multimedia search engine facilitates
access to open access documents provided by universities, museums, libraries, archives
and institutions all over the world. It indexes digital collections of every
kind (text documents, images, audio, video, animation, software, digitized
manuscripts, etc.).
(N18) The Sherpa RoMEO
colour system goes like this:
Green:
researcher can archive pre-print and post-print
Blue:
researcher can archive post-print (ie final draft post-refereeing)
Yellow: researcher can archive pre-print (ie
pre-refereeing)
White:
archiving not formally supported.
(N19) Impact is typically measured by the number of
times a paper is cited. Some studies suggest that open access increases impact,
according to OpCit, the Open
Citation project. Check for instance: Open Access to Research Increases Citation Impact, Hajjem, C., Gingras, Y., Brody, T., Carr, L. and Harnad, S. (2005).
(N20) There are now some 24,000 scientific
pair-reviewed journals and reviews worldwide.
(N21) LOLF
is: Loi Organique Relative aux Lois de
Finances (Institutional Act of the Finance
Law). The framework law on finance laws was passed in 2001, putting forward the need for
better indicators as a prerequisite to sound public spending. In the wake of
this law, Indicasciences was created with mission of analysing these
indicators and their relevancy. Leader in this field, is Thomson Scientific ISI
“web of knowledge” and Journal Citation Reports (JCR).
(N22) Last December, the 4 local universities and
regional institutions have decided to group into a European University of Brittany. Founders are: Université de Bretagne Occidentale, Université
de Bretagne-Sud, Université de Rennes 1, Université Rennes 2 Haute Bretagne,
Rennes Agrocampus, INSA Rennes, ENS Cachan–Kerlann and various
telecommunications schools.
(23) Irisatech club groups institutions and companies for technology transfers in the fields of computer sciences and telecommunications. Members benefit from services ranging from technological watch to training sessions on a variety of scientific topics.
---
----
This article in PDF format.
----
Other stories about Irisa/INRIA research center:
New e-Review in Computer Graphics
Branch Prediction under Scrutiny for Possible Security Flaw
Formal Verification and Conformance Testing for Reactive Systems
Analysis, Modelling and Simulation for Human Movements
Better Interoperability for Healthcare Information Systems
Peerple at Paris Linux Expo 07
Kerlabs in the Wake of Kerrighed
INRIA in Expert Group on Java Resource Management
Pushing Back Frontiers of Handwriting Recognition