Openness 'needs both political will and technical skill, fortunately ICRISAT has had both' an interview with Mr. M Madhan

Muthu Madhan is a practicing librarian in India for about 15 years and is currently the Manager for ICRISAT Library and Information Services. Before joining ICRISAT in 2010, he worked for M S Swaminathan Research Foundation and the National Institute of Technology Rourkela. Since 2004, he has been an advocate for open access movement in India.  Apart from open access, his other interests are open data,  open education, and open source software. 

The AIMS editorial team invited Mr. Madhan to share ideas and experiences on Open Access issues with the agricultural information management community.

Briefly outline Open Access related work done at ICRISAT?
ICRISAT has set up two institutional open access repositories -one for publications and the other for datasets to capture outputs produced by ICRISAT.Since May 2011, the publications repository has witnessed more than 240,000 downloads from more than 75 countries. ICRISAT has also conducted five workshops on open access for the Indian national agricultural research institutes and agricultural universities. As a result of these sensitization workshops, many agricultural research institutes in India have shown willingness to set up institutional OA repositories.On request we provided necessary technical help to Central Marine Fisheries Research Institute to set up their institutional repository. ICRISAT is in partnership with National Academy of Agricultural Research Management and the Directorate of Knowledge Management in Agriculture in open access advocacy initiatives. Finally, ICRISAT is one of the partners of the Agropedia initiative.

What would you say was the winning strategy to get this institutional commitment?
Discussions on open access to ICRISAT research output in the true sense began with an informal meeting between Subbiah Arunachalam and Alma Swan and a small group of our research managers in January 2006. Following that meeting the Director General (Dr. Willian Dar) of ICRISAT started a process of frequent internal discussions in the ‘committee of research leaders’ on open access to institutional research publications.  Culmination of the persistent and explicit efforts of the Director General was the 2009 open access mandate of ICRISAT. Dr V Balaji, the then global leader of Knowledge Sharing and Innovation program, who in fact executed the institutional repository system of ICRISAT,fondly explains this achievement in this report. Open Access needs both political will and technical skill,fortunately ICRISAT has had both.

What results have been witnessed after the implementation of ICRISAT Open Access Mandate?
In our case the mandate helps us in institutionalizing a brand new idea that is going to bring benefits to all in the research process. It just helps to infuse a new workflow in the scientific system. Mandate is just as simple as this. The institutional repository gives unique number for each publication/document uploaded in it. After the mandate, performance evaluation systems can ask the repository number along with the bibliographic details. Since 2011, it has become customary in ICRISAT that we submit the list of publications with “repository link” to each publication to the Governing Board.  Hence, mandate ensures 100% of ICRISAT publications come to repository and go open to all users and interested researchers.

In the developing countries within the agricultural domain it has been observed that a lot of research outputs have not been captured, what do you think could be done to capture this information?
Yes, this is a serious problem.  Recently, I and two other colleagues, looked at the papers published by Indian agricultural scientists.  We found, more than 70% of Indian agricultural research papers were published in journals published from India.  I am afraid that most of these journals do not even have web presence.  I am not sure these journals are subscribed by the other countries’ libraries.  Consequently, Indian agricultural research largely goes invisible outside India.India has only four institutional repositories related to agriculture and of these only two are functional. The case may be same for other developing countries.

What we need is a concerted and focused advocacy. I believe it is the responsibility of research institutes and universities to make sure their research outputs are in the public domain.  And I am sure interoperable institutional repository system is the best, cheapest and quickest way of achieving 100% open access to agricultural research output.  In this regard, and with adequate clarity, FAO and other international organizations should intensify their OA campaign and advocacy to persuade policy makers in national agricultural research systems.Also, I would say the OA campaign should focus the practicing researchers who have been indifferent for years.  

Most established repositories struggle to maintain a steady content growth and also most are not visible, What mechanism are in place to encourage the growth of repository contents?
It is a very good question.  Even if you have mandate in place, we must have a support system to help scientists to to acclimatize to this new workflow. In ICRISAT the library team has the responsibility of populating the repository.  We track ICRISAT publications vigilantly through publisher feeds.  We are alerted as and when a paper with ICRISAT affiliation is accepted by a journal.  We check the self-archiving policy of the particular journal publisher . Accordingly, we request (through email) the concerned author(s) the version that the publisher allows for archiving. In most cases, the journal publisher allows authors to archive post-print version in institutional repositories. At present, we follow this mediated archiving style.  We are doing this job meticulously. There is an improvement in scientists’ attitude; a few scientists send the post-print version of the papers before we alert them.  We are periodically sending the download statistics of the papers archived in the repository to the concerned scientists.

We have done some simple value-additions to the repository.One among them is that each document in the repository is linked with Google Scholar citations and that helps scientists to see who cites their paper through the repository system.  We are soon going to integrate the repository records with Web of Science which we have recently subscribed to. I would like to thank ICRISAT scientists for their cooperation in populating the repository

What activities related to Open Data are carried out by ICRISAT?
We have just created an interoperable data repository using Dataverse – open source software developed by Harvard University. The datasets repository is relatively a new initiative and we have just begun the campaign to populate it. We are testing a data harvester system and trying to perfect it.  You may see the test-bed here.We are customizing the system by tweaking codes to suit different types of datasets generated by our scientists.We are in the process of sensitizing our researchers on the need for bringing research datasets into public domain.We are carefully tailoring outreach programs to persuade researchers to archive datasets in the repository which otherwise they guard possessively.  We are trying to build a consensus on it.

In ICRISAT, the social science group has already made open the Village Dynamics Data that they have been gathering for over three decades. The applied genomics group, sticking to the community norms, has made open their plant genome sequencing datasets.  In recent times, publishers’ are increasingly demanding authors to submit datasets along with their papers.  But, research datasets are strictly institutional property.  I sincerely hope there will be sooner an institute-wide consensus on open data.