 Thank you that you participate in this presentation. We are delighted that so many attend at this webinar today. First of all, some organizational conditions. The webinar will take around about one hour. If you have questions during the presentation, please write it into the question and answer area on the button of your Zoom application. We take it here. Due to the high demand, we will collect questions during the presentation via the question and answer. Depending on the number of questions, we may only be able to handle a few at the end and help to answer as much as possible. And now, I'm excited about our speakers today. This is Christoph Brusinski from the OpenAPC project. He is a member of the DFG Funded Project Intact, and that starts four years ago. He has very much experience in this area of the APCs and OpenAPC field. My name is Andreas Czerniak. I'm the member of OpenAPC Advanced and responsible for the work package six two boards of scholarly comments and work at the University of Bielefeld in Germany. Before Christoph starts this presentation, I will give you a swift overview about OpenAir. OpenAir stands for Open Access Infrastructure for Research in Europe and started for 10 years and has around about 55 institutions across Europe. OpenAir promotes and provides support for access and open science across the European countries and has national open access desks in every country to support the national open access and open science infrastructure. And has also services for research and funders and projects and many more. OpenAir collaborate with CSEHUB, the Info Central, and many other organizations across Europe. So Christoph, the floor is yours. OK, thank you, Andreas, for this nice introduction. Can you hear me, loud and clear? All right. So just a moment, please. How do I switch it now? I don't know. The bad bit was too easy. OK, once again, hello, everybody. Again, thanks for this nice introduction by Andreas. Today I want to talk about the OpenAPC project. And actually, it's my first webinar I'm giving. I've talked a lot about OpenAPC already, but those were only very traditional face-to-face presentations. So let's see how this one plays out. So for some introduction, what is OpenAPC? To put it in a nutshell, OpenAPC is an open data project. And we are tasked with the collection, processing, and dissemination of data on APC expenditures. And I'm sure that most of you will be already familiar with the concept of APCs. But just for the sake of completeness, I'm also going a really small introduction into APCs. So an APC is an abbreviation for article processing charge. And basically, an APC is a one-time fee that has to be paid to a publisher when you want to publish an article, open access in a scientific journal. And paying an APC can either be mandatory in case it's a fully open access journals. In that case, APCs has to be paid for all articles. Or they can be optional in case of hybrid journals. Hybrid journals are journals which consist of a mixed set of fully open access and closed access articles. And when it's a hybrid journal, you are tasked with the decision if you want to pay an APC and have your article published OA, or you can choose to not pay an APC. In that case, your article will be published closed access. And just for a general rule of thumb, APCs usually range from 1,000 to 3,000 euros. The average value, which is reported to our project in old APC, is slightly below 2,000 euros. So and why are APCs an important topic? Well, the most important keyword, which always comes up when talking about this, is transition. At the moment we are right in the middle of a transition process. The scientific publishing landscape is in transition. It's slowly but steadily shifting to an open access world. And at the same time, the payments to publishers are also transforming because they are transforming from subscription fees to APC-based models. Because in the old world, payments to publishers usually were always based on subscriptions because the journals which were published had to be subscripted by libraries, by scientific libraries usually. And now this team of money is changing to an APC-based model. And the monitoring and steering and policymaking in this transition process usually requires solid data on APC payments. OK, now for some facts and numbers on open APC. Open APC has been established in 2014 here at Bielefeld University Library. And the first participants joining our project were German universities taking part in the open access publishing program of the DFG, the German Research Society. And what happened was that simply some people in our library thought, well, there are universities in Germany taking part in this publishing program. And they are all mandatory to report their funded articles back to the DFG anyway. So why not ask this universities if they would share their article lists also with us and established an open data project? And as it turned out, several of the universities agreed to this and they sent their data to us. And this was the start of the open APC project. This project then ran for about a year. And in 2015, we also received official funding by the DFG under the project name Intakt. And since 2018, we are being funded by the German Federal Ministry of Education and Research, the BMBF, under the project name COA. So what are our basic principles at the Open APC project? Our first and foremost principle is a maximum of transparency. So when we talk about open, to our mind, should not only apply to data, but also to things like coding, programming code, workflows, and the total project history. We'll come to that later how we managed to achieve that. We also employ a high degree of automation. We use many scripts and lots of programming code for things like data enrichment, for data testing, for data integration, and system building. Otherwise, this would be hardly infeasible because actually I'm the only person dealing with all the data processing. So if we wouldn't deploy this degree of automation, we wouldn't manage the whole data flow. And our third important principle is that we want to keep low entry barriers for new participants. For example, we don't have something like a formal membership. You don't have to sign any contracts or things like that. Again, we'll come to that point in a moment. So now something on our current state as of this month. Our current metadata collection consists of more than 75,000 articles with associated APC costs. They have been contributed by 231 institutions from 14 different countries. And the total reported APC amount is nearly 150 million euros. So even if the points on the first slide didn't really convince you why open APC is important, this number actually should because, well, that's a well-known fact. There's a whole load of money involved in the scientific publishing market. And lots of this money is now shifting to APC payments from subscription fees. Okay, let's get into the details, our data format and data processing. As I said, we like to keep it simple at open APC. So our whole data is reported and stored in simple table format. It's a simple CSV files. And our data format consists of 18 metadata fields, but only five of these fields are mandatory when reporting data to open APC. Because as we said, we like to keep low entry barriers. And all the remaining metadata is automatically imported from external sources. So here's a small overview on our metadata format. Actually what we see here is the first three entries from our whole data set. So the first three entries from our 75,000. And only these first five entries have to be reported by an institution if they want to report article APC data to us. So as we can see, this is, well, the institution's name, obviously, it's the University of Bamberg in Germany. The period the APC was paid in, the APC amount, of course, this is the most important metadata, and, of course, a DOI as an article identifier. And the information, it was a hybrid journal, the article was published in, or it was a full UA journal. Because that's the kind of point of information which we can't really automatically enrich. And almost all the other metadata fields we see here, are published from, are enriched from other external sources. For example, things like the publisher name or the journal title are obtained from CrossRef. We have the DOI. Also things like the ISSN identifiers or license identifiers or CrossRef data. Then we have the PubMed and PubMed Central identifiers. They are obviously imported from your PubMed Central. Also we are an API. So no manual work involved here. Then we have the ISSN L, that's the linking ISSN. That's a special identifier imported from the ISSN organization in Paris. Then we have the UT, that's a unique identifier for the Web of Science. So if the article is also in the Web of Science, we can import the UT identifier for it. And finally, we have a simple Boolean marker. If the journal in question is listed in the DOIJ, in the directory of Open Access Journals. So using this scheme of metadata enrichment has a lot of advantages. The first one, of course, I already mentioned that, it's much less work for participants because they only have really to report this five fields here and all the other fields are left up to us to fill them out. What's very important is that importing this metadata, for example, from Crossref, leads to unified publisher and journal designations. So we always have the same journal, full title and publisher names for the same journals, which we wouldn't have if our participating institutions would fill out these fields. And the third aspect is that corrections can be applied retroactively if we just run this enrichment process again for articles which are already in our database. Okay, let's talk a bit more about our infrastructure. What's very important is that all our data is maintained using Git. I don't know if you are familiar with Git. Git is a tool which is usually employed in software development, but as it turned out, Git can also be used for just for maintaining data. And so it turned out as a very useful tool for maintaining our datasets. And Git is also the advantage that those Git repositories can be published on services like GitHub. And we will just go there and take a look. This is our OpenAPC repository on GitHub. Now we can't see the URL, but if you have to believe me, we're on GitHub now. And as you can see here, first we have a readme file here which gives some general information about the data which is currently stored in OpenAPC. These are all the numbers we already saw on the slides before. Here are links to all the participating institutions and some statistics and plots, for example. And what's even more important is that all our data is actually stored and accessible here. We have this data folder here. Then we have subfolders for all participating institutions. And what's perhaps the most important part is this APC DE file here. The DE is remnant from the start of our project when it was really a German-only project. But nowadays it's really an international one. And we can also just take a quick look into this CSV file which has a decent size meanwhile. It's 17 megabytes. And so it takes some seconds for GitHub to actually display it. So don't let me down. Ah, there we are. And as we can see, well, it's a CSV file, as you would expect it. Very, very long, very lots of institutions in it. And as we can see, it conforms of the format I showed before. Again, we have these articles by Bamberg University which we have seen before on the slides. So now I have to see how I get back to my presentation because I can't see the headers. There we are. There we are. So the second piece of infrastructure I want to show is our project block because every time a participant contributes new data to OpenAPC, we write a new blog post in our project block where this data is analyzed in more detail. So we can see it here. For example, the University of Marburg recently joined OpenAPC, it was last Friday. And what we did is we generated half automatic blog post where the data contributed by the University of Marburg is analyzed in more detail with some tables, stats and plots. And back again. And our third and perhaps most well-known part of infrastructure is our tree website because we use interactive tree map graphics to visualize all our data. For example, let's go to the University of Marburg again. What we can see here is an interactive representation of the data contributed by the University of Marburg. And what's very appealing is that this tree map representation is interactive. For example, we can filter it because let's say we are only interested in ABCs paid in the year 2016. Then you can actuate a filter and tree map gets filtered down. And we can also delve into the data, so to say. So let's say we are only interested in articles published with the publisher, Springer Nature. Then we just click on this rectangle here and then we get down one level and now we are seeing all journals by Springer Nature where articles were published in. And we can do this process again. Let's just say we are interested in the journal Scientific Reports. One click and now we are at the article level. And we can even click on one of the DOIs here and then we will get to the landing page at the Springer Portal where we can actually read the Open Access article associated with this APC. Okay. End again. So this was for a short overview on the Open ABC project. And of course we're also doing a bit of advertising here because of course we're always on a lookout for more participants. So let's say you are an institution and you want to become part of Open ABC. What are your options? It's surprisingly easy, just as we promised. The first step, of course, you need data to contribute to us. So you would prepare an initial dataset according to our submission guidelines. What's important is that this dataset doesn't really need to be large. There are institutions participating which only have contributed about a dataset of five articles or so. So this doesn't really need to be a large table of articles and APC costs. Optionally, you would also create a small readme file which gives some background information on your institution and the background where your APC data originates from. Then you would also designate a contact person at your institutions with name and mail address. If there are questions regarding your data and then you would mail everything to our project email address or maybe see it in the e-mailerfield.de and that's it. We can look at the submission guidelines just really quick. Well, it's really what I already talked about in the presentation, manual requirements. There's the data schema you have to apply but it's really all the things I already mentioned, institution period, you, DOI and the Cybert, the mandatory fields. There's a definition of costs because it's not really that easy to define what an APC is. So there's an explanation what we expect this URL or how it should be interpreted. And here's this whole submission process described. So that's actually the most easy way of contributing data to us just saying an email and actually that's the way most institutions do it. I think about 95% of all institutions just mail their data to us in regular intervals. If you're feeling a bit more adventurous or you have very tech-heavy people working at your institutions, then you will also want to consider two alternative ways of submitting your data. The first one, this is a bit advanced is that you might initiate a pull request on our GitHub repository so we can adjust your data directly. It's actually just the same as an email but instead of sending your CSV or Excel file to us, you're directly contributing it to our GitHub repository. Again, our submission guidelines give a small guide how to do that. And if you are very, very, very advanced, there's a way, but this is, well, what you could do is you could modify your institutional repository so that APC course can be directly attached as metadata to article records. And I wasn't really sure if I would, I should really, I wasn't sure if to include this way of contribution into this talk because well, it's really difficult but if I understand it correctly, OpenAir is very keen about repositories. So I think I should mention it here. If you then tell us the relevant parameters, we can send a small harvesting bot to your institutional repository and then we can harvest your APC metadata regularly in regular intervals via the OIPMH interface. And what's very appealing about this is that it effectively reverses the flow of communications because you're no longer sending emails to us but just in reverse way, we are coming to you and harvesting the data from your repository. So that's very convenient for both of us because we are just having to operate a small harvesting bot and you are not under the obligation to send much emails any longer because it's all in your institutional repository. And we also have an example for that because the University of Regensburg here in Germany does exactly that. They're operating an E-Prince 2 server as the institutional repository. And if you're familiar with OIPMH, you might recognize this. There's a metadata prefix just for the open APC data sets and a metadata set. And we can also take a look at the OIE results and as you can see here, there are these metadata records corresponding to journal articles. And as you can see here, that's exactly the metadata we need at open APC. Institution period, Euro, and of course here the currency. That is actually all the data we need. So Regensburg actually has switched to this process of metadata harvesting from their institutional repository. So this is now all automated and they are no longer sending emails or initiating get pull requests. So as I told, let's go back on step. The first step of a contribution to OOP ABC obviously is to prepare an additional data set. So and that's actually what might become the most problematic because, well, as I said, providing OOP ABC with your data is easy and the most challenging step is probably just to obtain it in the first place. So why is that? Well, the data submitted to us usually originates from three different sources. The first source, those are those aggregated data collection from large funding organizations. For example, the Welcome Trust for Great Britain or the FWF for Austria. And well, funders usually, they require strict bookkeeping and reporting when paying for your APCs. So when you are publishing an article and you receive funding from, let's say, the Welcome Trust, then they usually require you to input all the relevant data on your article on some form. And well, it usually means that these funding organizations, they already have those large Excel sheets floating around that we can reuse at OOPM APC. And this also means that if authors at your institutions received funding from such an organization, then we might already have ingested that data just by receiving those aggregated collections. So point one, actually, that's where the most data receiving or getting to open APC comes from. But as an institution, actually you can't really do much about that. So point one, perhaps isn't the most important to you as an institution, much more important is point two, institutional publication funds. Because there are some institutions out there which operate local publication funds dedicated to FOSTA or a publishing within their institution. And speaking from experience, just as with the large funders, there's also probably collected data on funded articles that we can use because just like the big funders, like the Welcome Trust, also these smaller institutional publication funds, they also employ usually or should employ a strict bookkeeping. So it's also likely they again have a large, a bit smaller Excel sheet floating around. And there's a chance that we could actually use that data right from the beginning. And actually this might be the easiest way to provide open APC with APC data if you're an individual institution. And then there is the third source of APC data and well, that's what we like to call the gray area because that is actually everything else. So these are APCs which are not paid from any publication fund, but which are paid from other budgets. For example, by individual researchers, by research groups, by departments or by university faculties. And this is the largest of three sources and unfortunately it's also the hardest to dig into because in practice it requires to bring together two different worlds. So what do I mean by that? Well, the APC data itself, it's financial metadata and financial metadata, well, it's usually generated because a publisher is writing a bill for paid APC and this bill is then stored in accounting system, there are taxes paid for that. So we are now in the world of, well, financial accounting and on the other hand, everything else. For example, the DOI, which is our most important part of the metadata and things like the journal title, the journal hybrid status, that on the other hand is all bibliographical metadata and bibliographical metadata is handled usually by libraries and is stored in repositories that we just saw. So what we are seeing here is, well, you could call it a clash of cultures and it is in practice, it is really, really hard to obtain and map these two parts of information on each other. And unfortunately, we can't offer you really a catch all solution to that. There is no gold standard because if you want to dig into this data and try to map this financial metadata from your accounting system to bibliographical metadata stored in your repository, then you need a solution tailored to the situation at your institution. Fortunately, we can show an example how it can be done and our example is the situation as it is at the University of Milan in Italy. So the University of Milan is currently the only open APC participant from Italy and they do not operate a publication fund. If I'm not wrong, correctly, no Italian institution has a publication fund as to now. So there's no easily obtainable APC data for us. So none of these already mentioned floating around Excel sheets which we could obtain. So what the people at the University of Milan did is they developed a custom workflow to integrate APC data into their institutional repositories. So what did they do? They employed a two-step process. First, whenever an APC is paid, so there's a bill paid on APCs to a publisher, then the department paying the APCs can store additional metadata associated with this article in the institutional accounting system. At the moment, it's just the publisher and the author name, so nothing too fancy, but that's the first part of the process to later link this data to the article stored in the institutional repository. Because at the same time, the authors who are supposed to submit their articles to the institutional repository, they, on the other hand, can also fill out additional fields relating to article APCs. And these are three points of data, the hybrid status of the journal, the APC amount that was paid, and the currency. And what now happens is the people at the University of Milan, they verify all articles which are entered into the institutional repository on a regular interval. And if the submitting person turns out to be corresponding or first author of the article, this check is important because if the person is a corresponding or first author, it means, on the one hand, that the person usually is responsible for paying the APCs. And at the same time, it's an author affiliated with the University of Milan. So if that's the case, then the Milan people check if they have filled out the APC information in the step two, when entering the article into the institutional repository. And if this was not the case, they write an email to them and kindly ask them to do this now. And in the second step, they match and verify the information against the data entered into the accounting system in step one. So it's a kind of interlinked double checking process. And actually, it really seems to work because it's a good example of how to bridge this gap between those two cultures of, on the one hand, on the accounting world and on the other hand, of the library and repository world. Yeah, what they did is actually, they added some bibliographical data to the accounting system and on the other hand, some financial data to the repository system. And these two chunks of additional data added, they can then actually be used to match the articles on each other, the APC amount in the accounting system and the DOI and the other data in the bibliographical repository. And as we see, well, in general, if you're thinking about setting up such a workflow at your institution, an institutional repository is usually, to our mind, the best starting point to do this because you need some point where you start and in general, your institutional repository is the best bet because you need some point where you have the largest collections of article which are likely affiliated with your institutions and that point is usually your institutional repository. So it's probably the best starting point. So now for some applications of OpenAPC data. Well, as it's in general principle with all data, more insights can usually be obtained from the OpenAPC collection when it's combined with other data sets. So, and such endeavors have already been undertaken. I will show you three projects. We won't go into any detail here because we don't have the time for that unfortunately. But what has already been done, for example, one of our colleagues here at the National Open Access Contact Point here in Wielfeld, what she did is she examined the relationship between impact factors and APC amounts to measure if there's a relation between those two points. So for example, if journals with higher impact factors if they levy higher APC amounts, spoiler, they do. Another project, for example, was we obtained data directly from publisher. In this case, this was Frontiers. And with this publisher provided data, we were able to cross check against our OpenAPC data set to measure the effectiveness of these DFG publication funds here in Germany. And the third project was combining bibliomedical data from the Springer web portal, Springer link to estimate the effects of certain transitional agreements, the so-called Springer compact agreements. Again, I've added some links to all of this project and we will share the link to this presentation in the end. So if you're more interested in any of these sub-projects, you can easily click the links later and delve into it a bit more. But now for another thing, because since this is an OpenAPC, no, OpenAir, an OpenAir presentation, we also like to show you another combination between OpenAir and OpenAPC data. And this is something Andreas will now demonstrate you. So Andreas, you have the mic. Thank you very much, Christoph. So we will take over the screen. So, yes, what we did at OpenAir with OpenAPC data set, we use OpenAPC data set at data, Christoph present this and compare these information with the data sets in OpenAir and found around about 91%, which is 68,000 data sets in OpenAir, which is also in the APC. So as for this data set was completely all over time, the complete data, whole data set at OpenAPC, and the overall APC was around about 2000 euros per article. The total sum is 134 million euros. So as Christoph shows you, we create also a tree map for this. As you've seen before, you have the publishers, you can filter it out and so on. So you can select the date of the years and go into the depth of at least here for the publishers and go more and more in the details of which journal has how many APCs collected in this year and so on. So this is a cooperation between OpenAir and OpenAPC that we filter the open the data sets between this and both projects. And in detail, we do some more analytics. We find out that we have four major funders in the OpenAPC, OpenAir joining data sets with welcome trust with around about 10,000 articles and have around about 2,500 euros per article. As you can see, the next one is research concierge UK for OpenAPC and European Commission. And the fourth is with around about 3,000 the Austrian research science fund. And you see the different APCs there. So this is for project funders. What is the level of the funder programs? We can also take a look in this. I show you here the three major program of funders. Some FP7 has run about 6,000 articles and an APC of 2,100. And the next one is medical research concierge United Kingdom and so on. So we can compare these different programs of the funders and the APCs in these programs with the data of OpenAPC. And there's a lot of other funder programs here but these are the three. Going the next step deeper and we ask for the three major projects in these data sets and the data towards us. This is a human brain project from the European Commission has run about 18 articles but an APC was 4,000 euros. This is more than the others APCs I've shown you but gets mostly for medical have mostly more higher APC level as other journals. So the Neu-Methics was funded by European Commission welcome trust and research concierge UK. And the third one is AV Malar and also from European Commission and welcome trust. So what we do is actually we find out the way to present some APC information in the research community dashboards as an example or so as the explorer at OpenAir and more information with more analytics and evaluation can be done for other dashboards the funder dashboard and so on. And last but not least we are proposed that repository software developer include the APCs in the repository interface where APMH is one that Christoph here presented and a very nice thing to harvesting these information. Yeah, this is only the start of the collaboration between OpenAPC and OpenAir and we would to do more things automatically and enrich our data at OpenAir with OpenAPC data and presented to you. So this was from us of OpenAir and OpenAPC and then Christoph, that's your floor. Yeah, thank you, Andreas. Let's see if I managed to switch back. Ah, I'm getting better. All right. Again, hello everybody and unfortunately we are now nearing the end of our presentation. So just for short summary at the end. So what is what we do? OpenAPC collects and disseminates article-based APC data. We think that transparency on APCs is very important in the ongoing transition process to open access publishing. Actually, it's the first time that transparency is introduced into this whole academic publishing market because in the old world where subscription fees were usually the largest chunks of money being paid to publishers, there wasn't any transparency. So these subscription fees and how much money is paid that's usually still a large black area. So with OpenAPC it's really the first time that at least some transparency is introduced into this scientific publishing market. Well, as we saw contributing to our project is easy or at least should be easy. There's no red tape involved, no complicated data formats. You don't have to sign anything, no memberships or anything like that. There's also no need to submit large data sets. We don't have a minimum size. More is always better of course but every bit of data helps. If your institution operates an open access publishing fund it's likely that they already have some data we could use. So if you're working at an institution where such a fund is operating, well it would be nice if you could ask these people and make them aware. Perhaps they would be willing to share the data with us. And well, if that's not the case then you might think about establishing a workflow like the one we me landed and as we saw it's a bit complicated in the beginning but it might be worth the rewards because linking those article metadata to paid APCs will be useful in itself not only for us at the OpenAPC project. So then thanks for your attention all. And as I said, if you are interested in skimming through these slides again, for example, because you want to click some links or look at the subprojects I showed on the last slides, I have a link shortener URL here and you can find the presentation here and look at it for yourself. Thank you very much. Bye. Thank you, Christoph, for the nice, very good presentation. And now we, at the end of the presentation, we have some questions. So if you have questions about the presentations, please route your question into the Q&A area so we can collect this and we can answer this if we can. So first question was, since Patricia knows everything, have you tried to collect data from them or you think that there's no way they will be share money data? Sorry, I would love to answer this, but unfortunately, I can't see the questions here. Andreas has to find a way to show them to me. Ah, there it is. Thank you very much. OK, since publishers knows everything, have you tried to collect data from them or do you think there is no way they will share money data? Well, actually, that's a very interesting question. There are publishers who do that. For example, on the last slide, I'll show you this little subproject I talked about. And this was a project to estimate the efficiency of the German publication funds. And this project actually was conducted by a combining publisher data with our OpenAPC collection. And this was the Open Access Publisher Frontiers. So half of the answer to this question is, yes, they do. But usually, it's only the pure OA publishers who do that. For example, while Frontiers did it, we also had talks with, for example, Copernicus. They also were willing to share their data with us. I think the really big players, for example, like Elsevier or like Springer, I think they wouldn't be really willing to do that. So, well, the answer is yes and no. Andreas has it on his screen. So what's the next question? Great project and a very welcome one. Well, thank you. I may have missed it, but what would be very useful are data related to who has paid the APC. So even institutional name would assist in the investigations. Are there any plans to integrate the data sets from all the participants and include data regarding the origin of the APC payment? Yes, that is actually the case, because if you look back at our data format I presented, actually the first column indicates the institution which paid the APC data. And you can also see it in our tree representation. There is a section just to filter the tree maps for paid institutions. So if I understand this question correctly, yes. This is already included in our data set. OK. OK. Andreas has to mute his mic. Otherwise, OK, no echo anymore. If the journal is fully open access, not charging formal APC and funded by the government by a certain amount of money, could we consider average price of the published articles APC, which is not actually charging author institution funders, but taxpayers? Well, that's another interesting question. Actually, this one comes up in different flavors all the time. Actually, I haven't heard of government funded journals yet. But well, I'm always happy to learn. I wouldn't say that these are APCs which should be reported to our project. The reason is that at OpenAPC we are really interested in, well, what we like to say, pure APCs. So APCs which really conform to a transaction of money against open access publishing. And there are always some models which might be mapped into our project. For example, some years ago, there was this initiative by the World Society of Chemistry, RSC. What they did, they had developed a so-called voucher system. So as an institution, you could obtain some set of prepaid vouchers and then use these vouchers to publish articles open access, one per voucher. And then the question came up, couldn't we just divide the whole amount of money we paid for these vouchers against the articles published? Because that would then conform to a sort of APC amount. And we thought about that for a while. But we came to the conclusion, no, this isn't really an APC in the strict sense. So well, long story short, answer to this question is, no, this kind of data, which is only calculated in hindsight, because as you said, this would involve, again, some calculation. This is not APC data which should be integrated into our project. OK, any more questions? OK, but this is also transactions of money for all our publishing. Yeah, that's correct. But well, we have to keep the focus of open APC in mind. What many people sometimes confuse, or what they confuse open APC with is a monitor. And it is important to note that open APC is no open access monitor. So we are not trying to integrate a complete view of paid APCs to demonstrate a complete view of open APCs paid by an entity, be it a country or an institution or a research group. Because we are fully aware that this is impossible. We saw on the slides that there are three different races or three different sources where the APCs come from. And I also said that this gray area, this gray market of APCs, where APCs are paid by single researchers or publishing or research groups or faculties, that they actually make out the largest share of APCs out there. And we are fully aware that it would be impossible to grab this whole set of APCs just for one institution. And so open APC is no open access monitor. So actually the argument, but it's a transexual of money for oil publishing, that doesn't really apply. Because that would only apply if you were a monitor. But we are trying to establish a large sample of APCs paid. And that can be used in a second step, for example, to calculate average APCs, let's say, for a certain journal or for a publisher. And that number can be reused for other purposes, as we saw in the sub-project we established. OK, hopefully that answers these questions. Yeah, thanks for your question. So anything else? OK, we are a little bit over time, over an hour. So I think if you have questions, so write us an email to OpenAPC if it's directly for a question for OpenAPC. And how can you join OpenAPC? And for other things to OpenAPC and OpenAir, you please contact me. You see the contact details on the screen. And yes, thank you, Christoph, very much for this presentation, for this great OpenScience project and the practical introduction to APC. Thank you very much for your participating on this contributions. And we hope we have answered all questions. Yes, if you have further questions, contact us. So we close the webinar. And thank you all for participating and have a nice day. Thank you.