Big Open Science Podcast | S01E03: Sustaining and Innovating Open Research Infrastructures. SCIROS Workshop @UniLu

Transcript below

[00:00:04.240] 

Welcome to the third episode of the big Open Science podcast where we explore the present and the future of open science in social sciences humanities. I’m Tomasz Umerle, and today we are at the Centre for Contemporary and Digital History, C2DH at the University of Luxembourg. Our third episode is devoted to the open research infrastructures, specifically to the question of balancing, sustaining and innovating them. To discuss this we are joined by Tugce Karatas and Sofia Papastamkou from C2DH. Welcome Tugce and Sophia. Please say a few words about yourselves.

[00:00:42.370] 

So hi everyone. Thanks very much for inviting us to talk about these very important issues. My name is Tugce Karatas, I work at C2DH, the Center for Contemporary and Digital History here at the University of Luxembourg and I have a position as a research scientist. So it’s a little bit of the infrastructure, a little bit of the research and the technical part and I do day to day the help for the data management topics within the institution, but also contribute open science. And as a part of the digital research infrastructure team, I’m involving more in the digital part of our projects.

[00:01:33.000] 

Hello, and I am Sophia Papastamkou. I work at the C2DH as a postdoctoral researcher since 2021, and I am in charge of an editorial project ranked 2, which publishes open educational lessons on digital social criticism. I have a background of being a traditional historian of international relations and I have been further specialized to the application of digital technologies to history in the Ecole des Chartes in France and I have worked in France also as a research engineer in digital humanities for San Juan de Anacia Scientific.

[00:02:25.490] 

Thanks. Thanks guys for the introductions. So we are kicking off our episode with discussing challenges in sustaining open research infrastructures. Tugce, drawing from your experience in managing large historical data sets, what do you see as the main challenges in sustaining open research infrastructures in the long term?

[00:02:45.650] 

Yes, it’s a very important question. So we have as a team also we tackle quite a lot with this challenge because we are trying to sustain the infrastructure where we are also trying to innovate with the external projects and internal projects as well, because there are key challenges that are present in our infrastructure at the moment. So for example, the sustainability but also the long term funding as well as the data preservation strategies and also integrating the ethical and legal perspectives in the data governance. So these are the things that like very complex topics, but if I have to say in a very clear way that we need a structure that can support the sustainable tools and methodology, at the same time maintain and promote the Innovative solutions. So this is what we are trying to do within our team, but at the same time from the open research infrastructure perspective, that we need clear governance strategies, which involves the funding, which involves other stakeholders, which involves also the technologies, especially with the AI at the moment. So putting everything out as open science is also a little bit challenging at the moment for us.

[00:04:25.750] 

So what we try to do is that we respect, let’s say, the general best practice in the field, but at the same time try to, for example, integrate the open source tools and promote the open science practice for from the level that we show the example by doing the projects. So there are two different parts. So one of them that we are also trying to be the example, but at the same time we are trying to build the strategy. So at the moment it’s a little bit going hand to hand. But to reach this goal we need to start from looking at inside what we have. And, and if we have the solution of the repository, building our own infrastructure in the FAIR principles and all these standards, then we can go ahead and be more open in terms of the infrastructure. But at the moment we are having some challenges to integrate it in the institutional level.

[00:05:40.920] 

And Sophia, from your perspective, working on Rank two platform, which is an open educational resource, what other key aspects are crucial for sustaining open research infrastructures?

[00:05:52.040] 

Yeah, I will integrate some lessons that I also learned while I was serving to program historian. So we combined the two experiences. And so for me some key aspects are first, the institutional commitment and support to such initiatives, I very much believe to grassroots initiatives. So institutional support is not a necessary condition, but it can make things infinitely easier when it’s there. And at a certain point, even in the most successful community initiatives in a given moment, institutional support is always important. So this is one thing precisely. The existence of a community of practice is a very important aspect, because the commitment to use and to contribute to research infrastructures, actually user is necessary conditions for something to exist, something is sustainable by the use, the active use. Then some complementary aspects could be the use of low technology, minimal computing technologies, for example, because when a technology is simpler, it is more simpler to maintain it and de facto it becomes more easily sustainable. Also, low tech and minimal computing ensure easy access. Not everybody has the conditions to access very sophisticated infrastructure. Sometimes we have tendency to forget that outside of our privileged more or less universities, there are no particular good conditions of access, even for a decent connection of Internet.

[00:08:01.650] 

So an easily accessible project is used more easily and there we come again back to the user. And I think these are the Most important aspects, although I should not forget the importance of having good metadata so that our research data are easily discoverable. And I think that it this is one very much important aspect that open research infrastructures should ensure. And I will stop here. There are other things as well, but I think that these are the most important for me.

[00:08:45.390] 

So you both talked a little bit about the institutional support. So how much of this work towards sustaining different projects is led by existing policies where you are instructed to do certain things and how much you have to kind of build your own documentation around it. Can you say something about this?

[00:09:12.250] 

For me it’s 100% our own initiative because we don’t really have so much clear structure from like institutions because we are rather a young university and we are also rather a young interdisciplinary centre. So although we have our general structure as central infrastructure, we had to build everything on our own terms and needs. So I think for me it’s more like. So we have different of course the roles, but also different functionalities. So one of these functionality as a part of the DRI team, like the infrastructure team member is that like we are trying to of course also set the policies and effect. I mean like how do you say, go to the institutions, like European institutions, the university structure and innovation labs and all these different stakeholders to influence this, let’s say the open science principles. And that requires a lot of time and that requires also a lot of self questioning because like why do you do certain things but at the same time you feel like you are in power to influence these people. Because Luxembourg is a rather small country. So in maybe in another country it’s more structure and you kind of like find your way.

[00:10:55.040] 

But in Luxembourg it’s good or bad thing because once that you are in charge, you are in power of influencing this open science, let’s say discussions. But at the same time there is no structure. So it’s kind of like learning by doing so. It requires a lot of attention and being critical and like critical reflections of what you are we are doing. So we use a lot of our time to find our way in this sense.

[00:11:32.450] 

So sustaining open research infrastructures is often balanced with the need to innovate them to better address the needs of the changing research environment, especially in ssh. So Tugce your work, you are interested in the intersections of design, science and technology. How does your experience this area translate into thinking about innovations in infrastructures?

[00:11:55.790] 

Yes. So IMs. Yeah, so my career path is more or less like defined by the technology because I started my career as A product designer and then I change it to the digital curation and then I ended up like with the linked open data and all these like principles of let’s say digital data curation in a. In a way that makes sense for let’s say as you were mentioning before, like the principles, the strategies, but at the same time the tools and the methodologies. So I use a lot of design thinking, let’s say practice. But at the same time for me being a designer was also that literally the asking the question so what? So like if we are collecting all this data and if we are building for infrastructure in it’s like why we do these things and what do you want at the end of your project as an outcome? Because the technology and all these modular pipelines we can find always a solution with the recent development and AI is also affecting a lot of our decisions. But instead of using the tools I think we need to find a solution to use the structure as a methodology and understand that like why we are doing these things then rather than that like just stuck with one solution.

[00:13:31.150] 

So in our let’s say infrastructure we are also using that a lot of times this trainings and like peer to peer learning and rank two is one of those. But we need to also say that it’s not only me doing my research and me I know all the policies me, I know the best technology. But it’s more or less like we try to also share this and co design these processes. Otherwise it wouldn’t make sense for us as a team of diverse people, but at the same time as like running for 40, 50 projects at the same time. So we need to understand really that the core of the projects and react using this let’s say strategies like design thinking. I don’t know if I answered your question, but.

[00:14:33.580] 

So Sophia, based on Your work in Ranke 2 platform which aims to develop digital source criticism skills, what innovations in infrastructures can support the development of digital literacy among researchers in ssh?

[00:14:49.880] 

Yeah, I will mostly speak about historians because this is my background and my experience and historians have been and still remain quite conservative in these things. So the first innovation for historians is I think to really use digital technologies. Think there are research and there are a lot of resistance, although at a certain point we all use digital technologies. So this is why I think that for me what is very important is to be innovative in a simple way in design actually of the information systems that target historians and more broadly humanities and social science scholars. So think these systems based on the needs of research of people who are not particularly keen on using technology. And in this way this can really apply to actual research needs and initiate and get used people working in these environments. So I think this is a very important aspect and a second one that I can think of is to assure interoperability and portability of the research data. Because the most people are at ease in manipulating their data, I mean from taking simple notes, structuring them in a simple way in Excel sheets, then use data database, all these supposes manipulations and people that are not particularly used to doing this.

[00:16:51.340] 

If they use easy information systems, then it’s the system that guides them and it’s through using the system that they learn to do this. I think that this is behind. I can think of some software that enhances this kind of user, mainly Zotero to evoke an example or node code database management system where actually people can use them without being experts in either database management or structuring of data. But it structures simpler everyday research data we work with based on needs.

[00:17:40.680] 

On.

[00:17:41.480] 

The main needs of our research. And so it initiates smoothly people into using data. So this enhances people to obtain at least a basic level of data literacy. I think.

[00:17:59.060] 

So your centre attracts many people outside of your university, right? In Luxembourg many people come do their projects. But also one of the topics that you both talk about is how you influence the local researchers. So what is the, the balance your everyday life? Is your work more about kind of, you know, educating people that are already here or is it more about recruiting talents from outside Luxembourg to come and join you? How do you, how do you feel about this?

[00:18:34.540] 

I think it’s both like my. I. I also teach and at the bachelor course and I am also teaching the. That like Sofia was mentioning because I think we need to expose these tools to the people like whether it’s students or the general public, as Ranke 2 is doing, or our peers with the trainings. So I think there is a lot of community building going on at the C2D8. So it’s so like looking at the day to day jobs that we are having that we start with of course our own team. So I think we have the team exchange a lot. But at the same time all of our projects are also related to the Luxembourgish history. So we collaborate with a lot of external archives and also the BNL and National Archives, the museums on the linked open data. So we have this community also thanks to the national coordination of the DARIA infrastructure. So for us it’s like one of the things that we Carry on our DNA to learn. And you know, like we were discussing also before this, like knowledge exchange because it’s, it will, it will grow also with this time and it needs of course like building the community takes a time.

[00:20:22.420] 

But I think the impact at the end is much more big than just staying in the research and doing our research in the, in the university setting. But it’s also a part of the being interdisciplinary center. So we are very lucky in that sense.

[00:20:41.800] 

Yeah, yeah, we’re complete about our luck about working here because it’s. The staff is very International, the C2DH. So this, everybody over everyone. Each one of us brings their own network and we all combine our networks at a certain point. So this is very enriching for me for the project of Rankit2 because we need authors, reviewers, we work on the translatability of lessons since the beginning. So we conceive the lessons in order for them to be functional in more than one language and outside the English, which is very important. So this is one of the aspects. But there is also another aspect that is more local which I very much enjoyed this year because we all at a certain point teach. So this year I taught to two masters and I used our lessons on Rank two and other online materials. And so this was very enriching in the sense that there, there is less digital literacy. And it was a very interesting transfer because it kind of took us out of the centre where it’s not a bubble. But actually we are all more or less acquainted with digital technologies and so bring this knowledge to the level of master students and to sensitize them was very, very enriching and also gave back a feedback for enhancing and ameliorating our lessons.

[00:22:35.920] 

The work we are doing, I am doing on Rank two in the future.

[00:22:41.680] 

So going a little bit more broadly and looking at the future, what do you see as a key directions for the development of open research infrastructure in ssh? What are your expectations and hopes related to this area?

[00:22:59.880] 

Yes, I have a lot of hopes, let’s say, because I think with the advancement of the technology, I think we can also kind of find a way to delegate these mundane things that we are doing to maintain and sustain, but at the same time integrate these, especially with the artificial intelligence, to implement these new technologies to make the infrastructure more efficient and also kind of help us to reconsider the tools that we implemented five years ago to see that if it really makes sense to, for example, designate one person to one tool and then one person to do so, we are trying to create more like capacity building within our team, especially with these new technologies. But at the same time it gives us more flexibility to redesign and restart, structure our own, let’s say, infrastructure team. And also I think Sophia was very right to say that we are attracting more people and for example, for our data scientists, research scientists positions, they are like, we see that there is also a lot of, let’s say people coming to work in the university setting. So this interdisciplinary is quite important for us. But at the same time it’s a challenge because you need to find the balance of this development and the research at the same time.

[00:24:58.570] 

So I am really optimistic on these parts, but also a little bit of hesitant because as we were talking about the interoperability metadata and all the fundamentals, I think we don’t have so much control if we implement these technologies immediately. So we need to again just start from building the systems in the level of the clear data governance and then put the blogs on top like with the AI and the development of the tools and the softwares and also the open science. So it’s kind of like a good moment. We are trying to do step by step, but of course everything is mixed so you cannot really separate it. But at the end it’s a nice way to experiment and reflect on these topics. So I am quite optimistic about all these challenges.

[00:26:13.930] 

Yeah. As for me, I hope that open infrastructures will remain open first of all because political evolutions can be quite challenging in our times and things we take for granted may not be that granted. So yeah, I will not develop further on this, but I think there can be challenges in our times. A second element Tugce talked about data is integration with generative artificial intelligence, which I find is very promising for education and we are working this sense currently here. And the third element that is very important is the multilingualism. So just to enrich this environment which is dominated by the English, there is a lot of literature about this. I will not look further, but these are I think the most important aspects that I would like to remind. Thank you.

[00:27:27.000] 

Both of you touched a little bit on AI or a lot even in certain areas. So from my experience, I sometimes meet people, they are extremely enthusiastic. Right. They are excited about new possibilities of using AI to further their research, to speed things up. But on the other hand I also meet people and also teams that are very sceptical, very almost apocalyptic in their thinking. So I think that you are more on the positive side, but also balanced position. So what is your kind of view as A center and as a community, how do you see that? Do you also see those bubbles and what’s your reaction to that? And how do you think we should manage those or sometimes extreme positions?

[00:28:16.640] 

Yeah, it’s a good question because institutionally we were also like at the beginning with this, you know, is it a hype or is it something real? Then our reaction was like writing a manifesto and build a task force because we know the source criticism, we are the human humanist, digital humanist. And we know that how to, you know, like use these critical reflections and probably better than any other field because we know, we know the sources and everything. And what we did is we also tried to do a little experiment and see if it’s working. And it’s like what is the missing and what is not. So this is something I think we decided also that there are good things that we can do and use AI, but there are also the things that we can not use. It’s going hand in hand. In future we will also have assistant professor with us to define these challenges and give a little bit of the strategy. This was a reaction from our, our side and better like, other than that I think we can see also like with the, with many of our discussions and the groups that like we need to be very careful.

[00:29:49.860] 

I think we are not jumping on this topic easily. But at the same time there are people who are jumping and we don’t want to go behind. So I think university is also in a good practice because they already have their. Although we don’t have the policy or the guidelines, we. They are reacting in a way that like they are building another interdisciplinary center, digital ethics. So I think this is also the good step to talk, think about this consideration of the ethics and implement some kind of data governance policies. And since we are in the middle of Europe with this European AI act and everything, we are very responsive as well. So there are so many things going on to you know, like on one hand we go up and big. At the same time it takes us in the foundation that like we can, we can be more restricted on the. On the things.

[00:31:00.070] 

Yeah. As Tugce mentioned, we have a functional working group on artificial intelligence that was at hook formed in. At SDH and there is actually a big variety of work we do. Each of us actually works more or less by integrating generative artificial intelligence in research and in education. So just to talk about what’s my domain, we kind of started working on a small project, experimental project with how to, let’s say, take advantage of the fact that There are open educational resources in the public domain. So materials training materials of good quality, how to integrate them in LLMs and train the models in order to obtain a tutor in digital methods for humanists. So we started working on this. More will come in the coming months. Also this year we taught already we had more than one session, at least four sessions in the course of Introduction to Digital History. It’s a Master of European History. Here at the university, we integrated use of generative artificial intelligence. And we work with the students so that they are not afraid of this, so that they learn that it’s something that it is configurated by people and by machines.

[00:33:01.560] 

And they should not take for granted that the answers they obtain, but rather learn how to work with generative artificial intelligence. So we’re still in the beginning, but we continue this way.

[00:33:18.810] 

This has been a great discussion. Thanks to Che and Sophia for joining our podcast. It’s only in beginning, so keep up with our work. There’s more to come. Thanks.

[00:33:32.250] 

Thank you very much.

[00:33:34.330] 

Thank you. Stay connected and follow Skyros. For more insight, find us on our sciros.hypothesis.org 

Skyros project is made possible thanks to the support of the Polish National Agency for Academic Exchange under the Strategic Partnership Program. And remember, science is best when it’s open. See you next time on Big Open Science Podcast.

---

---