When did INRA first take steps towards making its information more 'open'? Why were these early steps taken - what was the key motivation?
The first steps were taken ten years ago by INRA Open access pioneers from our Animal Physiology & Livestock Systems Research Division and it mainly involved publications. They created the first eprint open archive in France registered in ROARMAP in 2004. They wanted an open and broad dissemination of their academic publications. In 2005, our CEO signed the Berlin Declaration on Open Access and it was the first political step in INRA’s involvement in the open access movement. Following this, we transformed our internal publications database into a OAI-PMH compliant repository. INRA is also the owner of 7 academic journals managed by international editorial boards and published by different publishers: Springer, Cambridge University Press, Biomed Central. Each of them has chosen its own open access system: gold with APC for those published by BMC or green for those published by Springer with a 1 year embargo (archive on http://hal.archives-ouvertes.fr/view_by_stamp.php?label=ARINRA)
How were these early steps taken? Were there barriers to these ideas within the organization? How were they overcome?
We started from the ground floor with pioneers and pilots. It has expanded with the executive board involvement. At the beginning, the researchers did not understand why we should have our own repository when their publications were available on Science Direct (Elsevier) or in the Web of Science (Thomson Reuters). We had to explain that it was not free access (around €1m for these 2 electronic packages!) and we had to find “key opinion leaders” in the INRA scientific community to promote open access. In 2013, 12% of INRA academic publications were published in open access journals, and 20% is stored in ProdInra.
INRA scientists are asked to reference their research outcomes (at large) in ProdInra. Depositing the full text is not mandatory as many outcomes are articles published in non open access journals. But we found that researchers also attach a document when depositing in ProdInra. Considering outcomes from 2010 to 2014, more than 40% of the records have a public attachment. Our involvement in European projects such as VOAR or Organic lingua has also been a strength for our dissemination policy.
For “data sharing” we are working with a group of 60 people (scientists, IT specialists, information managers) to define the different services for data management and data sharing. The experience shared with our partners in the different networks or initiatives like CIARD, GODAN and the Research Data Alliance are very important to help us in developing our policy and tools.
Can you explain the full scope of OA policies at INRA - do they extend beyond the ProdInra repository?
For the last 2 years, we have been working on open access to research data with a data sharing policy that aims to develop services to help researchers share their data in the most relevant way according to their disciplinary field or the strategic value of their results.
As a research institution we had many motivations in developing our open access or “open science” policy: i) to make our results more visible and accessible considering our scientific challenges (Climate Change, Sustainable Agriculture, Food and Nutrition …) ii) our main asset is “information” so we have to manage it and preserve it on a public information system which is not linked to private companies so that everywhere in the world, our information is accessible and specially for developing countries.
As you already understand, we are not only speaking about open access to publications but research outcomes at large and research data. The new paradigm of data-driven science is strategic for tackling our scientific challenges. We are building up an open science policy which is still “under construction” but with increasing results: OA articles in ProdInra, number of articles published by INRA scientists in OA Journals.
And from the "data point of view" we were one of the founders with FAO of the 2 groups " Agricultural data" and "wheat data interoperability” within Research Data Alliance which help us to address the relevant topics at an international level.
What have been the main results and benefits to INRA of ProdInra and other openness developments?
ProdInra has been a key tool for increasing scientists’ awareness of Open Access. Even if ProdInra addresses the green road only, the restrictions in providing a full access of the articles deposited in ProdInra have had an impact: INRA scientists now better understand that when they choose to publish in non open access journals, they transfer copyright to the publisher.
INRA produces scientific knowledge and research for economic and social innovation in the areas of food, agriculture and the environment. Its science is society-driven. It is important not to focus only on articles. ProdInra enables spreading research outcomes at large. Indeed, the targeted audience of INRA outcomes is scientific communities but also practitioners, professionals, teachers and students, policy makers, funders. The role of ProdInra in disseminating this research is very important.
What is planned for the future?
ProdInra is the first step in a more ambitious challenge to gather institutional information (activities, skills, outcomes). INRA expertise covers many fields and we are eager to provide an information system that highlights individual skills and activities.
The future development of our “open science” policy will consist in building a research data infrastructure that should have the following components:
- A service for minting persistent identifiers (DataCite DOIs) for our research data
- A catalogue to expose accessible collections of scientific data from INRA
- A common research data repository for the researchers who wish to move from a local file system storage to a shared repository
- Tools and guidelines for Linked Open Data publishing
In the meanwhile, we have started a project (Linked Open ProdInra) that aims to annotate and link out the content of ProdInra and make the results available as Linked Open Data. We want to build this with our partners. The different initiatives (Research Data Alliance or GODAN) and funding opportunities with H2020 will help us to shape this e-infrastructure for research in agriculture, food and environment.