 Hello, my name is Juno Remy and I'm very pleased to present at the CNI 4 meeting, the in-keep-it service that we've been building in the journey of Auschwitzland. I work as an assistant in Information Science at the University of Applied Science and Arts, and with three colleagues, Amélie Alwash, Arnaud-Goninna and Arnaud-Nichneida. We prepared this project briefing. This short presentation is divided into three parts. First of all, I will explain why we decided to create a supplementary service for allocating personal identifiers in Switzerland, then a few elements and the choice of identifier, namely ARC, and finally a few more slides that match the project. As you may be well aware, a PID is a long-lasting and bionic reference to digital resource. Technically, it consists of a server that is able to format or resolve identifiers to its corresponding web address. A PID has usually two parts. An identifier that can ensure the provenance of a resource and a location for the resource over time. Normally, even if the web address changes, normally a given PID should still resolve to its new location. If you take a look at the anatomy of the digital object identifier, there is a resolver service, a prefix, which is a number consisting of 10, which is the handle as part of the DOI namespace, and a number of the assigning body. And finally, the third part, the suffix, which identifies the resource. It is obvious that an identifier can only be persistent if there is a sustainable infrastructure or ecosystem in the background. And for this reason, I would like to give a shout-out to John Quincy, the person who created the Archival Resource Scheme, who in August 2018 wanted to debunk some of the myths surrounding PIDs. Any serious cultural orientation organization should consider these statements. As the Seoul University in the French-speaking Switzerland offering a curriculum in information science, we are very attached to the preservation and traceability of information. So these assertions are therefore in line with what we want to achieve. In Switzerland, we have several infrastructural services providing PIDs and mainly DOIs. There is a mixture of monopolized attribution. Sometimes those policies are contradictory and also works in progress and unclear legal issues, which leads to an unsatisfying and confusing situation. So that's why we wanted to build something else. And here you have in the map, for example, the ATR in Zurich is the main provider of DOI. They are a member of the data site, but it's quite costly and it's going to be even more next year, apparently. FORCE, which is for social sciences in Lausanne, they do provide DOI, but it's not for everything. The data lifecycle management TSM is more for long-term archival purposes, and not all data can actually be deposited there. Zinodo, that you may know, is from the CERN, the European Organization for Nuclear Research, but it's actually not legally in Switzerland, nor in France. It's extra territorial, so you have some legal issues if you deposit everything there. And then the DASH, the Data Service Center for the Humanities, so as the name says, is restraints to data in the humanities. So most of them use DOIs and then you have only DASH, which leverages the ARC scheme. The scheme that we actually are going to use as well, so I'm not going to talk about ARC, which is for us the best choice to satisfy scientific and cultural heritage sectors in Switzerland, in our opinion. The information in the following slides come mainly from the ARCs in the Open Initiative, which was in change, is named to ARC Alliance, and you have here two different URIs when you can find more information. So ARC is a scheme that was created in 2001 by the California Digital Library, and we can see that now over 600 organizations of the world are registered to assign ARC identifiers. These organizations are mostly based in North America and France, but we've also been seeing adoptions of different parts of the world, such as in South America and India. One of the very neat things with ARC is that it can be assigned to absolutely anything, from research data to bibliographic records, over a couple of times when you have a list of examples, but basically anything that would need an identifier can get an archival resource key. If you start assigning ARCs, you can indeed set up your own local resolver, which will also benefit from name to thing, which is a global resolver. So on the table you can see an example of the same identifier from the French National Library, and this manuscript not only points to the entire document, if you will, but to a particular page of this dissertation, which you have here on the slide. So to the folio number 29, an ARC can be subdivided into different parts. We've saw that with the structure of the DI, so the same thing with the ARC, but with more details. So here we can see on the top that there are three main things distinguished, the resolver service, the base, object name, and, and that's very important and relevant for us, the qualifiers, which we saw in the BNF example as well. If we go into a little more detail, we can see that the last two parts can also be divided. So you will generally find the ARC label, the name, assigning authority, number, that any organization can obtain as soon as they register. The assign name, which is like suffix in the, in the DOI, then you have a very fine level of generality where we can further differentiate elements such as pages, chapters, versioning, and even formats. You name it. You can divide the way you want your dataset. For instance, name to a thing, entity, it's a global ARC resolver, and it is actually for mapping names into things. So it knows where to root of a 900 of the types of an identifier. And when a resolution request comes in from the general public, entity looks up the identifiers and redirects the original link to forward the link. But to do this, it uses two different resolution patterns. To begin, entity tries to resolve according to information found in an individual stored identifier. And failing that entity tries to resolve according to any stored class rules based on the identifier type. So we have many advantages to use ARCs. First of all, it's free to register. Of course, it's not free to maintain an infrastructure, but still if you want to obtain a name, an assigning authority number, you have it for free. And it can be seen that ARCs can provide a very fine level of granularity through the use of parallel subdomains as well as qualifiers. So each organization is able to implement their own ARC policy and naming practices. And something that's quite neat in my opinion is its ability to fit with the TRIP-IF image API canonical URI syntax, which is not the case for most PIDs. With all these factors in mind, you can perhaps better understand why we have chosen ARCs as an identifier within the E-Inkipit project. So the project started in January this year is financed by the B5 program of Swiss universities. And we've been collaborating closely with the Swiss Institute of Bioinformatics, which is one of our partners, but also with the California Digital Library, EasyID, ARCs in the Open, now ARC Alliance. And thanks to them, we've been able to develop that infrastructure. So it will allow organizations to identify with a very fine level of granularity that can reflect complex object hierarchies, as well as being able to retain traceability versioning, sorry, or even information to the object's persistence. We've already started testing and within the next few weeks we'll start to actually assign ARC advice. And you can see in that diagram the different, so how we explain that to our customers. Once they have an ARC link, the end user can request it and how it's going to be forwarded also through name2t. So I'm not going to give a demo, but we basically have like EasyID, a user interface and an MPI. This is here the prototype. We've been reusing the EasyID components. We will of course have our own design and when we will upgrade the different components we want the EasyID infrastructure to benefit from that as well. So that was part, let's say, of the deal, that basically we would, what we do, any tools that we're going to create can be then later reused by the CDL. For the year 2021, again with funding from Swiss universities, we will continue to develop the infrastructure. So the main goal of the Incupids, Chris, that's a new name of the project, is to combine the basic service for assigning ARCs, which is here, Incupid 1 on that slide, with our current research information system, Chris, as well as a set of algorithm for search, disambiguation and interoperability. So more on that maybe next year and at a future CNI meeting, but that's what we're going to do next year just to develop the infrastructure and still and stabilize the ARC allocating service. So the creation of a service is never done online, and I would like to thank again all the people on this slide. So our partner from CIBS and non-nattery, Jean-Gobel, Luc Motin, Patrick Reuch, from the CDL, Isiandie, John Quincy, Kurt, Evelton, Maya Gould, Greg Janne from UC Santa Barbara, as well as a partner, a financial partner, Autiodean and ICRAF, and of course, ARCs in the Open ARC Alliance for very interesting conversations. So thank you very much for your attention, and if you have any questions or remarks, I will happily answer them.