 Session discussions and if you don't mind I'll start with group number one funders and data policy and in our group we had basically two types of people research funders and also people supporting data and research infrastructures in the countries and There was an interest from Danish Research Council in Introducing open data policy and reusing some of the warnings that mark presented today Then There was a comment from research data storage infrastructure in Australia about data value checklist Something that would be really interesting to look at because what they did in Australia as they come up with their requirements from disciplines from discipline communities and They check whether they are service provider can Met all these requirements. So that would be interesting to compare Australian data value checklist and UK data value checklist Then there was a comment That it's really cheaper to keep Data then to try to recreate them because sometimes it's really more expensive to recreate data There was a question from Nordic Council of Ministers If you have a situation when Some data were generated from research funded project Plus there were some data sets brought by research like external data sets then can you require Open availability of that part of data set that was brought into project And it's really the answer is that you should probably approaches case by case in general. Why would researcher try to find to try to hide someone if He or she is good then he or she should be happy to share But there is a question whether it's only access or reuse only so in this case case by case Negotiation should go and Then if research projects have been funded by multiple funders then Whether data policy of one funder would be applied and Recommendation was to have this data policy that would be applied to research funded in whole or in the pot That's that's the best approach Mark mentioned that Codata is preparing a toolkit for the funders how funders can come up with open data policy So I think that would be really useful resource Then there was a comment from UK repository net What is the role for Institutional repositories in this whole data process and also it's somehow related to a comment and question from Spanish foundation for signing technology for science and technology and It looks like institutional repositories will be good places to Handle some Data related to certain experiment for example or to certain publication, but speaking about data pools like really data sets That were created by pulling Data from different sources. It's probably something that only data centers can do because that's kind of added value of data centers and for institutional repositories, it's very important to have clear and proper Metadata information so if some if some data are deposited in repositories, they should be really well documented to because They say the goal is to have them reused later and you need proper documentation Then there was a question from Swedish national data service how do you handle data management plans after the project so if there was no data management plan and the project is over and possibly if Data has some value and there is a risk of loss of those data Then maybe a data management plans could be funded poster project But in general, it should be still institutional responsibility Institution should have internal audit or what data sets are available. How they are managed Then We had discussion about identifiers that Again related to data in repositories or in data centers That it would be good to have Researchers identifiers then it would be good if researchers would acknowledge funding sources It's kind of common in the US to acknowledge your fund there It's not so common in Europe and it would be good to introduce as good practice And then when you have research and identifiers you have those avoid identifiers. You have DOI's then you can pull together Projects information publications and data and There was a comment from University of Botswana whether Because we were talking about some national data management plans whether there might be some International approaches to data management plans in certain areas and there might be some opportunities means this. Thank you And I'm passing over to Inger okay So we talked about institutional policies and We saw that in most institutions. There is not yet a policy and not even The project of getting a policy because it's not easy to do Now what I especially got out of the session that we had is Many institutions ask themselves what other steps that we have to do before getting such data management policy So First of all be aware of the hierarchy. Who do you have to involve? Who do you need to talk to to get everyone in this and and to Want to to join the project and be happy to do so So be aware of it hierarchy identify academic champions not People really just focusing on their one niche a Subject but more academic champions in Reaching out throughout different disciplines different sorts of material so that if you work with such champions You get a broad overview of what are the needs of researchers because that's something that you have to take with you What are the needs of the researchers? Try to get into their processes. So they start an extra administrative burden That's something that you have to take with you in your mind and in the development of such a data management plan Be sure not to contradict other policies. It has been said before so Be careful about that so identify which policies are in place in your country in most used Funders that your researchers engage with so that you know what the policies are there Certainly align what happens in align with what happens in other external Repositories because it was also identified as a problem that many smaller institutions and even rather big institutions Don't have the ability don't have the infrastructure the money the budgets To get into a big data management infrastructure. So to keep all your Data safe and everything but what you can do is develop a data management plan that guides researchers throughout the process of research Failing redoing things and data to keep in an archive in in an infrastructure so that you can at least guide them and give procedures and advice and things like that also What do they have to? Deposit in that repository what do they have to keep? It has been said before a checklist is needed a Basic checklist. What is the data that we have to keep? What do you mean with primary research data? So that is Certainly important. There's a few of the key things we talked about If I look at it from an open-air open-air plus View how can we help? People how can we help institutions to develop such a management plan the data management plan because that is what we want to do is a Help or engage in in the projects in in the talks discussions about putting up such a checklist and and try to Get in Help with that. That's what I want to say Also guidelines and developing policies. I already mentioned some things But if we can make a sort of guide to develop Management data management policy so that you can see what happens in other institutions. What are the things that we have to look at? Also, if you want to do a survey in advance, what kind of questions do you need to ask? What do you want to know before starting such a management plan? So we do have a few things in open-air plus that we can do to help institutions get into such a data management policy and So that is something that we can do. So that is what I get back from take with me from this talk this breakout group So I will report back from group three researchers and publishers. We started off with the two questions Do researchers want to share is data publication a reality and All all over the discussion typically the examples you You get to the table of from life sciences or environmental sciences and Overall, there's a recognized need for for further Cases in other other fields where you can demonstrate the value of things and we touched several questions in this in this field like In a typical experience could be of a researcher that He or she does not felt acknowledged for the the effort which has been put into the collection of data It's it's then at a later stage where people are get acknowledged that like you have some methods You have the tools and these people are cited, but the others are not cited So this is then something which is a practice, which is of course a bad thing and is in is a hindrance in between sharing so a question which was We discussed say quality issues like peer review what kind of principles I mean, this is something which is emerging which is not there yet under which kind of principles you should Proceed when you peer review research data So and it's typically the ones who use the data which are candidates for reviewing the data and And there are then various degrees of openness how you handle the the the peer review and of course how you handle the the sharing of data which we touched already in the in the talks and and one Discussion a longer discussion was about what makes researchers Nervous about sharing data. What are the issues on their side? It's it's typically of course quality assurance quality issues and Well the point of promotion that can always mention of their of their Work and then about I mean, they're not always confident about the value of that data They they have some have the attitude like well, my data is not good enough for sharing or whatever And we felt it is well, it would be good to endorse this principle and give them more Confidence in this this area and leave it then to the to the field What is what is done with the with research data when shared? so overall we should aim for Changing the habits of researchers such that Sharing is not something which happens at the beginning of the project and at the end But it should be a process which you touch on a daily base like do it five minutes and things are easy so Yeah, and Publishers also have a role in this area like they can help to in enforcing standards We had Theodora from plus in our group which we shared some Experiences also like when it comes to quality. It's important that you have have the original For example pictures you some researchers use them, but you well then you go back. Do they Still have the original part the original data yeah, and No, one one of the last points was about data citation He's not should not be the only thing we are looking at also alternative matrix Which are not there because there's not so much a critical mess of of these kind of things But like for publications, this is of course something which we should look at so one one point of summary like I Already mentioned this like Aim for key key ex I mean aim for good examples, which which can be used in this field to To educate researchers and the and and push push things and also a list of trusted Databases where where to share data is something which is Obvious need so that's it what I can If there's anything else from the group, please I'm presented the results of That the discussion that we have of the technical on the technical discussion We were a small group of people of six people. I think just one outside of the open air Consortium the other ones are people that are already involved, but it was interesting to to To know about from exactly from that that person that is not currently working on on open air about an initiative in Finland of constructing a national data catalog that will collect research data from all disciplines and it will have a A mix of meta that only records and the meta that and that is and data sets And of course, we immediately and polimately start thinking that we could use that for for harvesting and for connecting to to to finish data sets then we we have a small discussion on change of ideas on on that on the open air data model and Especially we have we spend some time Talking about the the fact that on the data model and services of open air we we are using Trust levels for for entities and and for relations and that those trust levels can be used Not only for the automatic relations that so the relations that can be established by By our computational methods, but also for man-made relations as Aryan Pointed out on on his presentation on an asset publication that in some cases we will have man-made relations and the data model of open air can also use that levels of trust to to apply to to to that and And and also and then we discussed that the issue of the need for get off of Guarantee or at least every information about the long-term preservation and long-term availability of Datasets and other objects on an S publications for an S publications to be trustable. We need to to to To be sure for hopefully to be sure that all the the components are Will be there in five years or ten years or one hundred years or at least to be to know what parts of the of the of that An S publication will be will we'll have a long life and Eventually what parts will probably ever shorter life and so with we also talk about that We can also use kind of levels of trust for that and from from those levels of trust We can then develop services at user interface level so we can show users what kind of relation and what kind of of preservation It's associated with that object and also we can for instance develop some automatic Services for instance for low trust preservation objects we can then periodically sent our Robots to check if the object is still available and if not just drop it from from the relations with with the publication so I think that's the the main issues that we have discussed it I don't know if Aryan or Paolo want to add something else Okay Right in two minutes. We had a very small group and a lot of what's just been said I'm not going to repeat because we covered quite a lot of those points In terms of what relates to open a plus There was a question about metadata and whether or not there were licensing restrictions for the metadata itself not just for Databases, so we may need to think about that also technically how we're going to link We should get sort of the case studies with our scientific communities out there Published as soon as possible in terms of how we're linking going to link up from our information space to to external data sets and databases I Think that There were two research issues flagged up that the credit for researchers should be really made very clear in terms of sharing data and Usability was really really crucial The role of the library and the sort of inertia that comes from the library sector and fear perhaps launching into sort of the data management world Was perhaps a reason why libraries don't have data management policies perhaps they're not best suited And that needs to be tackled and also universities are fairly difficult places. They're very disparate with different Units and divisions and there's a lot of politics going on So one one person said that it's quite hard Politically to to put forward a data management policy that everybody agrees on And That's really the those are the main points that that we went through because there are lots of in-between Discussions, but as we're running out of time. Those are the points. I'd like to tell you so I think I'd like to bring the Workshop to a close. Thank you very much I'd like to spank thank the speakers very much because I think that they were all excellent really good level of Talks and a really good range. Thank you for all of your participation in these breakout groups Please go and enjoy a nice lunch I'd like to also thank the open-air team because this has all been a team effort to pull this off today So thank you all very much and last but not least and order bibb as well I hope that this has stimulated some interest within your regions for the open-air partners You could go back and start to discuss about data management policies how you might start Thinking about preparing one yourselves at the very least. I hope this has interested you in this topic Thank you very much and Enjoy the afternoon