All entries for Monday 08 August 2011
August 08, 2011
IRIOS Workshop – Part Two: Comment and Workshops
Writing about web page http://www.irios.sunderland.ac.uk/index.cfm/2011/8/1/IRIOS-Workshop-Parellel-Sessions
One thing I took away from the workshop session was that both systems ROP and IRIOS were doing the right things and going in the right directions but weren't quite there yet. A big concern to me as an IR manager (and as a former Metadata Librarian) was that the IRIOS system creates yet more unique identifiers (see later in this entry for further discussion of unique IDs). Also automation of the project linking to outputs can't come fast enough, especially for services like WRAP where we spend a not inconsiderable amount of time tracking down funding information from the papers. However we could also benefit from taking information from systems such as this, which tie the recording of information about outputs much more closely to the money, which is always a motivator for people to get data entered correctly!
I think it is telling that more and more of these 'proof of concept' services are being developed using the CERIF dataformat (after R4R I'm looking forward to hearing about the MICE project early next month) but the trick with a standard is that it is only a standard if everyone is using it. I don't think we are quite there yet, I think this coming REF has been such an uncertain process so far that I think there is a lot more chance of CERIF being the main deposit format in the next REF. (If I'm still here for the next REF I'll have to reflect back on this and see if I was right!)
The afternoon of the work shop was taken up with a number of workshop discussions on a range of topics, below are a few of the notes I took in the two discussions I took part in. To see the full run down of all of the discussions please see the link above.
Universal Researcher IDs (URID)
It was generally accepted by all in the discussion that unique IDs for things, be they projects, outputs, researchers or objects were a good idea in terms of data transfer and exchange. They must be a good idea as there are so many different ones you can have (in the course of the discussion we mentioned more than eight current projects to create URIDs). Things are much easier to link together if they all bear a single identifier. However when it comes to people the added issue of data protection rears its head and can potentially hamper any form of identification if it is 'assigned' to the person. A way round this was suggested to allow people to sign up to identifiers, thus allowing those who wish to opt out to do so. Ethically the best route perhaps but unless a single service was designated we could end up with a system similar to the one we have now where everyone is signing up, but not using a whole array for services. The size of the problem is the size of the current academic community and global in scope. Some of the characteristics of URIDs we came up with were they just be; unique (and semantic free - previously mentioned privacy issues), have a single place that assigns them, have a sustainable authority file, not be tied to a role. One current service in place that fulfils many of the above criteria is the UUID service, however this falls down in that there is no register of assigned IDs so people can apply for multiple IDs if they forget them (and lets face it the likely hood of remembering a 128 number is kind of low) ... and we're back in the same situation again. I'm not sure there is a single perfect solution to this problem, though my life would be easier if there was!
This was a free form discussion that covered the REF, REF preparations and 'Life after the REF' in various guises. HEFCE are currently tendering for the data to be used in the REF at the moment, needless to say the two services bidding are the expected two, Thomson Reuters and Scopus, but HEFCE will only be buying one lot of data. Bibliometrics were touched upon in relation to the REF, is it better to have two people select a really highly cited paper or choose two lower cited papers? Discussions on the HESA data, checking the data once it comes back from HESA, possibilities of mapping the future HESA data to the REF UoA for long term benchmarking rather than a single point hat goes out of date very quickly. Do people's CRIS systems really hold all of the data required for a return? What are the differences between the impact as measured/requested by HEFCE and the Impact measured by RCUK? Selection policy and training, the possibility of sector wide training, possible best practise mentioned in the idea to train a small core group of people who would handle all of the enquires centrally. Would it be possible for institutions to get the facilities data on a yearly basis rather than just before the REF and then have to try and chase people who may not remember/have left to try and verify the data?
One interesting comment from the discussion was the news that NERC, at least, has seen a big increase in the number of grant applications including a direct cost for Open Access funding. Interesting particularly is that there had been a number of comments made to me that researchers didn't want to do that are they feared making their grant application too expensive.
All in all the day was very interesting for me as an introduction to a 'world beyond publications' (as I was attending both for myself and for a member of our Research Support Services department) and as an indication of what we need to do to go forward.
IRIOS Workshop – Part One: The Presentations
Writing about web page http://www.irios.sunderland.ac.uk/index.cfm/2011/7/28/IRIOS-Workshop-Presentations
The IRIOS (Integrated Research Input and Output System) workshop at the JISC Headquarters was designed to demonstrate the preliminary look of a system designed to take information from the RUCKL funders on Grant awards and combine it with the information from University's IRs and CRIS systems. The event was attended by research managers, representatives of four RCUK funders and repository managers and all of the presentations can be seen at the link above.
The event kicked off with a video presentation from Josh Brown of JISC discussion the RIM (Research Information Management) programme of projects. One interesting statistic was that it is estimated the £85mill/year is spent on submitting grant proposals and administering awards. Once again the savings that can be realised with the use of the CERIF data format was raised and the point that REF submissions can be made to HEFCE in CERIF was highlighted as a sign of the growing importance of the standard. IRIOS was highlighted as a step towards a more integrated national system of data management. Josh closed with the news of a further JISC funding call to investigate further uses of CERIF due to be announced soon.
Simon Kerridge (Sunderland) was up next to discuss the landscape and background of the project and the need for interoperability and joined up thinking between a number of different University departments if we are to make the most of an increasingly competitive environment. He also spoke of the ways in which IRIOS might feed into the RMAS (Research Management and Administration System) project further enhancing the cloud based system. Simon finished by touching on the challenges (research data management and unique researcher IDs anyone) and opportunities for the future (esp. the JISC funding call).
Gerry Lawson (NERC) was up next with a whistle stop tour round the RCUK 'Grants on the Web' (GoW) systems. Starting with a stern warning that if the funders and institutions don't find a way to match up the data held by both parties commercial services will find a way to fill the gap (for example Elsevier's SciVal is already starting this process), thus putting both parties back into the situation where we have to buy our own data back. Other products are also making a start on this process, as can be seen in the UK PubMed Central's grant lookup tool. Gerry made the vital point that all of the information is available but linking it is going to take work. The RCUK 'Grants of the Web' system is a start in this process as it brings together all of the grants by all RCUK funders in a single system. The individual research councils then use this centralised data to populate their individual GoW interfaces with each interface being set up to the specifications of the individual research councils. With one exception, AHRC, grant data about individual funded PhDs is not included in the GoW systems due to the RCUK preference for handling funded PhDs through their network of Doctoral Training Centres. Gerry closed saying there was a real desire from the RCUK to start linking outputs with funding grants (and expanding into research data and impact measures) especially in relation to monitoring compliance with Open Access mandates. Challenges still remained; a need for a common export format (CERIF); authority files for people, projects, institutions; the issue of department structures within institutions changing over time etc.
Dale Heenan (ESRC), ably assisted by Darren Hunter (EPSRC), discussed the RCUK 'Research Outcomes Project' (ROP). The project was based on the ESRC's Research Catalogue (running on Microsoft Zentity 2.0) extended to meet the needs of the four pilot councils, AHRC, BBSRC, EPSRC, ESRC. (Worth noting that MRC and STFC use the e-Val system). The ROP system is designed to create an evidence base to demonstrate the economic impact of funded research and also designed to attempt to reduce duplication of effort. Upload of data can come from a range of stakeholders, grant holders (PIs or their nominated Co-Is), institutions, IRs etc. and can cover individual items or bulk uploads. Management Information is provided using Microsoft Reporting Services. Future challenges for the system include ways to automate the deposit of research outputs, development/adoption of standards such as CERIF, ways to pull data from external services like Web of Knowledge, PubMed Scopus etc.
The main presentation for the day is of the IRIOS demonstrator by Kevin Ginty and Paul Cranner (Sunderland). The IRIOS project is a 'proof of concept' demonstrator of a GoW like service using the CERIF dataformat and is based on the 'Universities for the North East' project tracking system (CRM). One feature of the service is that four levels of access (hidden, summary, read only, write) can be assigned to three distinct groups (global, individual, groups of users) allowing a fine level of dynamic control over the data contained in the system. All grants and publications have a unique ID that is automatically generated by the system and any edits mad in the current system do not feed back to the system that originated the data. Currently the system is only accepting manual linking of grant to output but there are plans to look into automation of this process. In the future it might be possible to import data from larger databases like Web of Knowledge but information gathered by the research councils indicates that only 40% of outputs are correctly attributed to the grant that funded the research.
If you would like to try the demonstrator version of the IRIOS system details on how to login can be found here.
Comments on the presentations and information on the workshops is to follow in part two.